Journal of Adolescent Health 46 (2010) 444–450

Original article

Peer Stressors and Gender Differences in Adolescents’ Mental Health:

The TRAILS Study

Martin P. Bakker, M.Sc.a,b,*, Johan Ormel, Ph.D.a,b,c, Frank C. Verhulst, Ph.D.d,and Albertine J. Oldehinkel, Ph.D.a,b,c,d

aInterdisciplinary Center for Psychiatric Epidemiology, University Medical Center Groningen, University of Groningen, Groningen, The NetherlandsbGraduate School of Health Sciences (SHARE), University Medical Center Groningen, University of Groningen, Groningen, The Netherlands

cGraduate School of Behavioral and Cognitive Neurosciences (BCN), University Medical Center Groningen, University of Groningen,Groningen, The Netherlands

dDepartment of Child and Adolescent Psychiatry, Erasmus University Medical Center – Sophia Children’s Hospital Rotterdam, Rotterdam, The Netherlands

Manuscript received April 17, 2009; manuscript accepted October 8, 2009

Purpose: This study tested two hypotheses about gender-specific mental health effects of peer

*Address correspo

Center for Psychiatric

CC72, P.O. Box 30.00

E-mail address: m

1054-139X/10/$ – see

doi:10.1016/j.jadoheal

stressors during early adolescence: (1) boys and girls are sensitive to different types of peer stressors,

and (2) peer stress is associated with different mental health problems in boys and girls.

Methods: These two hypotheses were tested in a prospective large population cohort of 2,084 Dutch

young adolescents. Internalizing and externalizing problems were measured at baseline and follow-up,

whereas stressful life events in the period between baseline and follow-up were measured retrospec-

tively at follow-up. We performed the analyses with two types of peer stressors; victimization at school

and relationship losses.

Results: Relationship losses were more strongly associated with internalizing and externalizing prob-

lems in girls than boys, supporting the first hypothesis. Peer victimization at school was also associated

with both types of mental health problems, but equally strong in boys and girls.

Conclusions: Peer stress is unlikely to be associated with different mental health problems in boys

and girls. Instead, boys and girls are more likely to be susceptible to different types of peer stressors.

� 2010 Society for Adolescent Medicine. All rights reserved.

Keywords: Peer stressors; Mental health; Adolescents; Gender

During adolescence, the peer group becomes the primary

context for socialization and emotional experience [1]. These

new bonds not only broaden and enrich the world of adoles-

cents but also lead to higher interpersonal demands [2]. Peer

group affiliation, social status, and intimate close relation-

ships become more important for adolescents [3,4], which

in turn increases the risk of loss, rejection, and conflict.

Consistent with this increased risk for interpersonal difficul-

ties, peer victimization and stressful events involving close

friendships and romantic relationships have been found to

be major forms of peer stress[1], which are detrimental to

adolescents’ mental health [5–7]. For reasons to be explained

ndence to: Martin P. Bakker, M.Sc., Interdisciplinary

Epidemiology, University Medical Center Groningen,

1, 9700 RB Groningen, The Netherlands.

.p.bakker@med.umcg.nl

front matter � 2010 Society for Adolescent Medicine. All

th.2009.10.002

hereafter, we hypothesized that peer victimization and rela-

tionship losses may lead to gender differences in the mental

health of adolescents [1,7].

Adolescents’ mental health is likely to be affected by loss

of belongingness that occurs when social bonds are threat-

ened, broken, or refused [8–10]. Belongingness is a funda-

mental human goal, which can be characterized as the need

to form and maintain social bonds [8]. Although this goal

is assumed to be universally present in human beings, boys

and girls might seek belongingness in different social spheres

[11]. Specifically, boys are likely to pursue belongingness in

the broader peer group by competing for a good social posi-

tion within a status hierarchy, whereas girls may pursue

belongingness more in dyadic close relationships [1,7,11–

14]. We propose that peer victimization at school is prototyp-

ical for a low social position in the peer group [15,16],

whereas relationship losses (e.g., loss of friendships,

rights reserved.

M.P. Bakker et al. / Journal of Adolescent Health 46 (2010) 444–450 445

romantic break ups) are prototypical for stress in dyadic peer

relationships. On the basis of possible gender differences in

pursuit of belongingness, mental health in boys might be

particularly affected by peer victimization at school, while

mental health in girls is more affected by relationship losses.

Gender differences in mental health might also be due to

a differential orientation in the expression of psychological

distress [17–19]. Specifically, sensitivity to peer stress in girls

may be reflected in inward emotional responses, whereas in

boys it may be more reflected in outward-directed behaviors

[7]. In other words, gender differences in internalizing and

externalizing problems could be explained by a gender-

specific sensitivity to particular peer stressors or by gender-

specific expressions of psychological distress [7].

To the best of our knowledge, these two explanations for

gender differences in mental health have never been tested

within a single study. To recapitulate, we expect that peer

victimization is more strongly associated with internalizing

and externalizing problems in boys, and relationship losses

are more strongly associated with both types of mental health

problems in girls. Alternatively, we hypothesize that both

peer victimization and relationship losses are more strongly

associated with internalizing problems in girls and external-

izing problems in boys. This study aims to add to the current

published data by testing both hypotheses in a prospective

large population cohort of Dutch young adolescents.

Methods

Sample

Subjects were participants of the ‘‘TRacking Adolescents’

Individual Lives Survey’’ (TRAILS), a prospective cohort

study of Dutch adolescents, aimed at explaining the develop-

ment of mental health from preadolescence into adulthood.

The TRAILS study was approved by the Central Committee

on Research Involving Human Subjects (Dutch CCMO).

Sample selection involved five municipalities in the North

of the Netherlands, including both urban and rural areas.

The five municipalities were requested to give names and

addresses of all inhabitants born between October 1, 1989,

and September 30, 1990 (first two municipalities: mean

age¼ 11.29 years; SD¼ 0.52; range, 10.0–12.0) or October

1, 1990, and September 30, 1991 (last three municipalities:

mean age¼ 10.72 years; SD¼ 0.37; range, 10.0–11.5).

Two birth cohorts were used to minimize the age range during

the initial assessment. A detailed description of the sampling

procedure and methods is provided by Huisman et al [20].

Of all the children who were approached (N¼ 3,145),

6.7% (n¼ 211) were excluded because of mental or physical

incapability or language problems. Of the remaining 2,934

children, 76.0% (N¼ 2,230; mean age¼ 11.09; SD¼ 0.56;

range, 10.0–12.0; 50.8% girls) were enrolled in the study

(i.e., both child and parent agreed to participate). Of the

2,230 baseline participants, 96.4% (N¼ 2,149) participated

in the first follow-up (mean age, 13.56; SD, 0.53; range,

12.0–15.0; 51.0% girls), held 2–3 years after baseline (T1)

(mean number of months, 29.44; SD, 5.37; range, 16.69–

48.06).

The present study is based on data from both the baseline

(T1) and follow-up (T2) assessment wave. Written informed

consent was obtained from the parents and from the adoles-

cents themselves at both assessment waves. During these

waves, questionnaires were filled out by the adolescents, their

parents, and their teachers. The adolescents filled out their

questionnaires at school, in the classroom, under the supervi-

sion of one or more TRAILS assistants. Both the parent(s)

and adolescents received a small gift after participating in

the study of which they had no prior knowledge (parents

got a pen with the TRAILS logo and the adolescents a gift

certificate of V10 euro). Responders and nonresponders did

not significantly differ in levels of problems behaviors or

on sociodemographic variables [20]. For the present study,

valid data on both stressful life events and internalizing and

externalizing problems were available for 2,084 adolescents

(97.0 % of the follow-up sample).

Measures

Internalizing and externalizing problems. Two broad

domains of mental health problems were included in this

study: internalizing and externalizing problems. These prob-

lems were assessed at baseline (T1) and follow-up (T2) with

the parent-rated Child Behavior Checklist (CBCL) [21], the

Youth Self Report (YSR) [22], and the Teacher Checklist

of Psychopathology (TCP).

The TCP was developed by TRAILS to reduce the respon-

dent burden for teachers, as each had multiple participants to

report on. The TCP is composed of descriptions of problem

behaviors similar to Achenbachs’ Teacher Report Form

(TRF) [23]. The TRF is the teacher-report version of the

CBCL and YSR [23]. The TRF was designed to identify

the same eight syndromes as the CBCL and YSR. The

three syndromes, anxious/depressed, withdrawn/depressed,

and somatic complaints, fall under the larger grouping of

internalizing. The two syndromes, aggressive behavior and

rule-breaking behavior, are grouped under the group of exter-

nalizing. The remaining three syndromes, social problems,

thought problems and attention problems, do not fall under

either of the two mentioned groupings:. Thus, the TCP yields

the same syndrome and domain scales as the TRF, CBCL, and

YSR but based on (sets of) single vignettes rather than sets of

items. For example, the vignette for withdrawn/depressed

syndrome is: ‘‘The adolescent wants to be alone rather than

to have company. He/she is withdrawn and has little contact

with others. The adolescent doesn’t show initiative or shows

a lack of energy’’; the vignette for aggressive behavior is:

‘‘The adolescent is conflictuous and challenges others. He/

she bullies others and physically attacks them. The adolescent

has an explosive and unpredictable nature. He/she gets easily

frustrated by others and frequently uses abusive language.’’

Response options for each description of the TCP ranged

from 0 (not applicable) to 4 (very clearly or frequently

M.P. Bakker et al. / Journal of Adolescent Health 46 (2010) 444–450446

applicable). The TCP vignettes correlated around .60 with the

full TRF syndrome scales filled out by a small sample of

teachers (unpublished internal report, available on request).

We created broadband scales of internalizing (anxious/

depressed, withdrawn/depressed, and somatic complaints)

and externalizing (aggressive behaviors and rule-breaking

behaviors) problems. The validity of these scales has been

documented [21–23] and reiterated in a Dutch sample

[24,25]. In our sample, the reliability statistics for the baseline

sample (T1) were as follows: CBCL-internalizing (32 items,

a¼ .85), CBCL-externalizing (35 items, a¼ .90); YSR-inter-

nalizing (31 items, a¼ .87), YSR-externalizing (32 items,

a¼ .85); TCP-internalizing (3 vignettes, a¼ .71), TCP-

externalizing (2 vignettes, a¼ .78). For the first follow-up

sample (T2), the reliability statistics were as follows:

CBCL-internalizing (31 items, a¼ .86), CBCL-externalizing

(35 items, a¼ .90); YSR-internalizing (31 items, a¼ .88),

YSR-externalizing (32 items, a¼ .86); TCP-Internalizing (3

vignettes, a¼ .71), TCP-externalizing (2 vignettes, a¼ .80).

Reports from different sources are needed to reduce rater

bias in the prediction of mental health problems and provide

better estimates of diagnosis than those based on a single

source [26,27]. For this reason, we computed a combined

measure of mental health problems, using the scores given

by the adolescents, parents, and teachers. To place the

same weight on information from different informants, the

scores on CBCL, YSR, and TCP were first standardized to

mean zero and standard deviation one (z-scores) before aver-

aging over informants. When data of one informant were

missing or unreliable (for internalizing: CBCL: T1 n¼ 157;

T2 n¼ 225; YSR: T1 n¼ 41; T2 n¼ 17; TCP: T1 n¼ 281;

T2 n¼ 601. For externalizing: CBCL: T1 n¼ 148; T2

n¼ 203; YSR: T1 n¼ 32; T2 n¼ 2; TCP: T1 n¼ 279;

T2 n¼ 590), the composite score was based on the other

informants. The composite scores of internalizing problems

and externalizing problems were subsequently standardized

to mean zero and standard deviation one (z-scores).

Stressful life events. Stressful life events were assessed retro-

spectively with a self-report questionnaire at follow-up (T2),

including 20 stressful life events. The items had a yes/no

Table 1

Prevalence rates of the specific peer stressors during early adolescence, by gender

Girls (n¼ 1,069)

N (experienced event)

Loss of close frienda 160

Romantic breakupa 228

Victim of physical violencea,b 22

Victim of negative gossipa,b 256

Victim of bullyinga,b 219

Victim of sexual harrassmenta,b,c 46

a Stressful life events are dummy variables (0, not experienced; 1, experienced)b Victimization experienced at school.c Includes both verbal and physical forms of harassment.

* p < .05.

format to indicate whether the event had occurred in the

last 2 years (between baseline [T1] and follow-up [T2]).

For each stressful event, the level of experienced severity

was also asked, which the adolescents could rate as 0, not

unpleasant; 1, somewhat unpleasant; 2, rather unpleasant;

or 3, very unpleasant. We used this measure to exclude events

that were not experienced as unpleasant.

Of the list of 20 stressful life events, we selected those

peer-related events that were prototypical for either peer

victimization at school or relationship losses. Peer victimiza-

tion at school includes physical, verbal, and psychological

forms of degrading actions by peers [5,28]. We selected

the following prototypical events for this category: victim

of physical violence at school, victim of sexual harassment

at school (both verbal and physical), victim of bullying

at school, and victim of negative gossip at school. The cate-

gory of relationship losses included the loss of a close friend-

and romantic breakup. The prevalence of the specific stressful

life events within each category is given separately for boys

and girls in Table 1. Numbers of stressful events within these

two categories were used as variables in the analyses.

Statistical analyses

Gender differences in peer stressors and mental health

problems were examined using t-tests. Associations between

the variables were tested for boys and girls separately by

Pearson correlations. Multiple linear regression analyses

were used to test gender differences in the effects of peer

victimization and relationship losses on internalizing and

externalizing problems. All continuous variables were stan-

dardized to M¼ 0 and SD¼ 1, and gender was dummy

coded (0¼ girls, 1¼ boys), so as to ease the interpretation

of the interaction effects. For each mental health outcome,

we tested the main and interaction effects of peer victimiza-

tion at school and gender, as well as the main and interaction

effects of relationship losses and gender. The interaction

effects were only maintained in the model if significant.

We adjusted for pre-event mental health problems because

some events might in part reflect prior manifestations of

psychiatric symptoms [29].

Boys (n¼ 1,015) c2 test

N (experienced event)

40 c2 ¼ 72.723, p < .001*

152 c2 ¼ 14.171, p < .001*

39 c2 ¼ 5.974, p < .01*

129 c2 ¼ 43.534, p < .001*

176 c2¼ 3.236, p¼ .072

37 c2¼ 1.301, p¼ .254

.

M.P. Bakker et al. / Journal of Adolescent Health 46 (2010) 444–450 447

To provide an impression of the effect size and facilitate

the interpretation of the interaction effect, we wrote out

multiple equations using simple slope analysis [30]. In these

analyses, low and high levels of the predictor indicate one

standard deviation below and above the mean, respectively,

while holding all other variables to their sample means.

Results

Descriptive statistics

The rate of friendship loss was higher in girls than in boys.

Girls also reported more romantic breakups than boys. Being

the victim of bullying and sexual harassment did not differ

significantly between the genders. However, boys reported

more physical violence and girls reported more negative

gossip (Table 1). Girls reported significantly more peer

victimization at school and relationship losses than boys,

although the mean differences were small. Girls had more

internalizing problems, whereas boys had more externalizing

problems. Gender differences in internalizing problems

increased, whereas gender differences in externalizing prob-

lems decreased between T1 and T2 (Table 2).

The two types of peer stressors were weakly to moderately

correlated with each other. In both boys and girls, peer

victimization at school showed a stronger correlation with

internalizing than with externalizing problems, whereas rela-

tionship losses correlated about equally strong with both

types of mental health problems (Table 3)

Peer stressors and gender differences in mental health

For neither internalizing problems nor externalizing prob-

lems were the interactions between peer victimization at

school and gender statistically significant (internalizing

problems: B¼�.04, P¼ .220; externalizing problems:

B¼�.04; P¼ .317). In contrast, the interaction effects of

relationship losses and gender were significant for both inter-

nalizing and externalizing problems (Table 4).

Simple slope analyses [30] revealed that the effect of

relationship losses on internalizing and externalizing

Table 2

Mean (standard deviations) of peer stressors and mental health, by gender

Girls (n¼ 1,069)

Mean (SD) N (experienced events)a

Peer victimizationb .51 (.77) 1,069 (383)

Relationship lossesb .36 (.57) 1,069 (336)

T2 internalizingd .16 (.85) 1,069

T2 externalizingd �.06 (.77) 1,069

T1 internalizingd .03 (.72) 1,069

T1 externalizingd �.18 (.64) 1,069

a The total number of girls (or boys) who experienced peer victimization or relab Sumscore.c Test statistic adjusted for unequal variances.d Scores based on the mean standardized parent, adolescent and teacher reports.

problems was stronger in girls than in boys (internalizing

problems– boys: B¼ .05; t¼ 2.55; p < .05; girls: B¼ .16;

t¼ 4.35; p < .05; externalizing problems– boys: B¼ .06;

t¼ 3.20; p < .05; girls: B¼ .14; t¼ 3.80; p < .05). These

results are represented in Figures 1 and 2.

Discussion

Main findings

In this study, we examined possible gender differences in

the effect of peer victimization at school and peer-related

relationship losses on two types of mental health problems,

that is, internalizing and externalizing problems. We hypoth-

esized that boys’ mental health would be more affected by

peer victimization at school, whereas the mental health of

girls would be more affected by relationship losses. In

addition, we hypothesized that peer victimization at school

and relationship losses were more strongly associated with

internalizing problems in girls and externalizing problems

in boys.

The results showed that relationship losses were associ-

ated with more internalizing and externalizing problems in

girls than in boys, which provides further support for the

notion that girls are more affected by changes in dyadic

peer relationships than boys [1,7]. However, this sample

was relatively young to meaningfully examine romantic rela-

tionship loss. Nevertheless, the prevalence of romantic

breakup in our sample was sufficient to investigate gender

effects for both genders. But different effects may have

been observed for an older sample, where presumably dating

behavior will be more frequent.

Like relationship losses, peer victimization at school was

associated with both internalizing and externalizing prob-

lems. Contrary to our hypotheses, we found no gender

difference in mental health problems as a result of peer

victimization, which might suggest that being rejected by

peers at school is a universal stressor during early adoles-

cence. Nevertheless, the experience of victimization at school

might be stressful to boys and girls for different reasons.

Tentatively, peer victimization threatens pursuit of a good

Boys (n¼ 1,015) Difference t-test

Mean (SD) N (experienced events)a

.37 (.70) 1015 (268) t¼ 4.19c, p < .001

.19 (.43) 1015 (178) t¼ 7.88c, p < .001

�.15 (.71) 1015 t¼ 9.01c, p < .001

.08 (.82) 1015 t¼�3.92, p < .001

�.05 (.72) 1015 t¼ 2.63, p < .01

.18 (.82) 1015 t¼�11.39c, p < .001

tionship losses.

Table 3

Correlations between peer stressors and mental health, in girls (above the diagonal) and boys (below the diagonal)

Peer victimization Relationship losses T2 internalizing T2 externalizing T1 internalizing T1 externalizing

Peer victimizationa .27 (<.001) .31 (<.001) .18 (<.001) .20 (<.001) .17 (<.001)

Relationship lossesa .26 (<.001) .20 (<.001) .21 (<.001) .05 (0.10) .09 (<.01)

T2 internalizingb .33 (<.001) .12 (<.001) .45 (<.001) .50 (<.001) .26 (<.001)

T2 externalizingb .12 (<.001) .10 (<.01) .37 (<.001) .30 (<.001) .57 (<.001)

T1 internalizingb .27 (<.001) .12 (<.001) .55 (<.001) .23 (<.001) .49 (<.001)

T1 externalizingb .13 (<.001) .08 (<.01) .21 (<.001) .57 (<.001) .43 (<.001)

a Sumscore.b Scores based on the mean standardized parent, adolescent and teacher reports.

M.P. Bakker et al. / Journal of Adolescent Health 46 (2010) 444–450448

status position by boys within the peer group, whereas for

girls peer victimization poses constraints to establish affilia-

tive ties at school.

Previous research provides mixed support for gender

differences in mental health because of peer victimiza-

tion[1,31], which could be due to heterogeneity within this

category of events. Boys are assumed to be more affected

by overt victimization, whereas girls experience more mental

health problems caused by relational victimization [1,31].

We performed post hoc analyses to examine this possibility.

Gender differences in mental health were tested separately

for being a victim of physical violence (overt victimization)

and negative gossip (relational victimization) at school.

The interaction with gender was not significant (results are

available on request) in either of the regression models, sug-

gesting that our combined measure of peer victimization did

not obscure gender-specific pathways toward mental health

problems.

The results for both relationship losses and peer victimiza-

tion at school did not support the hypothesis that peer stress is

associated with different mental health problems in boys and

girls. Hence, our findings suggest that gender per se is

unlikely to result in different reactions to the same peer

Table 4

Main and interaction effects of peer stressors and gender on mental health

problems

T2 internalizing

problemsa

T2 externalizing

problemsa

Bb (p) Bb (p)

Model 1c,d

Peer victimizatione .20 (<.001) .07 (<.001)

Gender (1¼ boys)f �.29 (<.001) �.10 (<.01)

Model 2d

Relationship lossese .16 (<.001) .14 (<.001)

Gender (1¼ boys)f �.29 (<.001) �.08(<.05)

Relationship losses 3 gender �.11 (<.01) �.08(<.05)

a Regression analyses for T2 internalizing problems are adjusted for T1

internalizing problems; Regression analyses for T2 externalizing problems

are adjusted for T1 externalizing problems.b Unstandardized regression coefficients.c Main effects are presented because interaction effects between peer

victimization at school and gender were not significant.d All continuous variables in the model were standardized to M¼ 0;

SD¼ 1.e Sumscore.f Dummy coded variable (0¼ girls, 1¼ boys).

stressor; rather boys and girls are more likely to react to

different peer stressors.

In additional, girls reported more peer victimization at

school than boys. One well-documented explanation is that,

in general, girls are more likely to report interpersonal stress

and conflicts than boys [32,33]. It is therefore possible that

girls are more willing to admit peer victimization than boys

[33]. Another explanation for this result could relate to

gender differences in the onset of puberty during early

adolescence. According to the social misfit hypothesis, early

maturers are perceived as social deviants by same-age peers

[34]. As a result, early maturers are more likely to be targeted

for victimization by peers [34]. In general, girls tend to expe-

rience puberty at an earlier age than boys [35]. Thus, more

girls than boys were likely to have been early maturers in

our relatively young sample. This might also possibly explain

why we found that girls reported more peer victimization at

school than boys during early adolescence.

Limitations and strengths

A limitation of this study is that the life events were

measured retrospectively, without regard to contextual infor-

mation, and were based on self-reports [36]. Retrospective

self-reports are likely to be susceptible to recall bias. For

example, people with mental health problems have been sug-

gested to over-report the severity as well as the number of

Figure 1. Interaction between relationship losses and gender on internalizing

problems.

Figure 2. Interaction between relationship losses and gender on external-

izing problems.

M.P. Bakker et al. / Journal of Adolescent Health 46 (2010) 444–450 449

stressful life events [36,37]. Tentatively, recall bias in self-

reports about peer victimization at school and relationship

losses might be more pronounced in severity ratings than in

occurrence ratings of these stressful life events [38]. To

reduce possible state-related biases, we only used the number

of stressful life events in the analyses, not their reported

severity. Important assets of this study are the size and repre-

sentativeness of our sample, the use of multiple informants,

and multiple outcome measures. The longitudinal nature of

our study made it possible to adjust for pre-event mental

health problems as well.

Implications and Conclusion

To the best of our knowledge, this study was the first to

examine two plausible explanations for gender differences

in mental health because of peer stressors in a large prospec-

tive population cohort of Dutch young adolescents.

An important result of this study is that gender per se does

not predispose adolescents to either internalizing or external-

izing problems in response to peer stress. Whether adoles-

cents react to stressful events by developing internalizing

or externalizing problems does not seem to be strongly

related to gender. Previous research suggests that differences

in temperamental traits partly explain the risk for either inter-

nalizing or externalizing problems in adolescents [39]. For

example, adolescents with low effortful control are more at

risk to experience externalizing problems than internalizing

problems, whereas this pattern is reversed for fearful adoles-

cents [39]. Thus, low effortful control and fearfulness, rather

than gender, might in part account for a possible differential

expression of mental health problems in response to peer

stress. However, this conjecture awaits future research.

Another noteworthy result of this study is that rejection by

peers at school is an important stressor for both genders

during early adolescence. Furthermore, girls reported more

peer victimization than boys. Taken together, girls might

be particularly at risk for peer victimization during this devel-

opmental period. Our results therefore provide additional

support to the emerging view that girls’ experience with

victimization at school deserves much closer scrutiny from

researchers, parents, and teachers [40].

Acknowledgments

This research is part of the TRacking Adolescents’ Indi-

vidual Lives Survey (TRAILS). Participating centers of

TRAILS include various departments of the University

Medical Center and University of Groningen, the Erasmus

University Medical Center Rotterdam, the University of

Utrecht, the Radboud Medical Center Nijmegen, and the

Trimbos Institute, all in the Netherlands. Principal investiga-

tors are Professor Dr. J. Ormel (University Medical Center

Groningen) and Professor Dr. F.C. Verhulst (Erasmus

University Medical Center). TRAILS has been financially

supported by various grants from the Netherlands Organiza-

tion for Scientific Research NWO (Medical Research Council

program grant GB-MW 940-38-011; ZonMW Brainpower

grant 100-001-004; ZonMw Risk Behavior and Dependence

grants 60-60600-98-018 and 60-60600-97-118; ZonMw

Culture and Health grant 261-98-710; Social Sciences

Council medium-sized investment grants GB-MaGW 480-

01-006 and GB-MaGW 480-07-001; Social Sciences Council

project grants GB-MaGW 457-03-018, GB-MaGW 452-04-

314, and GB-MaGW 452-06-004; NWO large-sized invest-

ment grant 175.010.2003.005); the Sophia Foundation for

Medical Research (projects 301 and 393), the Dutch Ministry

of Justice (WODC), and the participating universities. We are

grateful to all adolescents, their parents, and their teachers

who participated in this research, and to everyone who

worked on this project and made it possible.

References

[1] Rose AJ, Rudolph KD. A review of sex differences in peer relationship

processes: Potential trade-offs for the emotional and behavioral devel-

opment of girls and boys. Psychol Bull 2006;132:98–131.

[2] Compas BE, Wagner BM. Psychosocial stress during adolescence:

Intrapersonal and interpersonal processes. In: Colten ME, Gore S,

eds. Adolescent Stress: Causes and Consequences. New York, NY:

Walter de Gruyter Inc, 1991:67–85.

[3] La Greca AM, Harrison M. Adolescent peer relations, friendships, and

romantic relationships: Do they predict social anxiety and depression?

J Clin Child Adolesc Psychol 2005;34:49–61.

[4] Pellegrini AD, Long JD. A longitudinal study of bullying, dominance,

and victimization during the transition from primary school through

secondary school. Br J Dev Psychol 2002;20:259–80.

[5] Graham S, Bellmore AD. Peer victimization and mental health during

early adolescence. Theory Pract 2007;46:138–46.

[6] Monroe SM, Rohde P, Seeley JR, Lewinsohn PM. Life events and

depression in adolescence: Relationship loss as a prospective risk factor

for first onset of major depressive disorder. J Abnorm Psychol 1999;

108:606–14.

[7] Rudolph KD. Gender differences in emotional responses to interper-

sonal stress during adolescence. J Adolesc Health 2002;30S:3–13.

M.P. Bakker et al. / Journal of Adolescent Health 46 (2010) 444–450450

[8] Baumeister RF, Leary MR. The need to belong: Desire for interpersonal

attachments as a fundamental human motivation. Psychol Bull 1995;

117:497–529.

[9] Bowlby J. Attachment. Attachment and Loss, Volume 1. New York,

NY: Basic Books, 1969.

[10] Bowlby J. Separation Anxiety and Anger. Attachment and Loss,

Volume 2. New York, NY: Basic Books, 1973.

[11] Baumeister RF, Sommer KL. What do men want? Gender differences

and two spheres of belongingness: Comments on Cross and Madson.

Psychol Bull 1997;1997(122):38–44.

[12] Cyranowski JM, Frank E, Young E, Shear K. Adolescent onset of the

gender difference in lifetime rates of major depression: A theoretical

model. Arch Gen Psychiatry 2000;57:21–7.

[13] Geary DC, Byrd-Graven J, Hoard MK, et al. Evolution and develop-

ment of boys’ social behavior. Dev Rev 2003;23:444–70.

[14] Troisi A. Gender differences in vulnerability to social stress: A

Darwinian perspective. Physiol Behav 2001;73:443–9.

[15] Salmivalli C, Huttunen A, Lagerspetz KMJ. Peer networks and

bullying in school. Scand J Psychol 1997;38:305–12.

[16] Veenstra R, Lindenberg SM, Zijlstra BJH, et al. The dyadic nature of

bullying and victimization: Testing a dual-perspective theory. Child

Dev 2007;78:1843–54.

[17] Gjerde PF, Block J, Block JH. Depressive symptoms and personality

during late adolescence: Gender differences in the externalization-inter-

nalization of symptom expression. J Abnorm Psychol 1988;97:475–86.

[18] Gjerde PF. Alternative pathways to chronic depressive symptoms in

young adults: Gender differences in developmental trajectories. Child

Dev 1995;66:1277–300.

[19] Leadbeater BJ, Blatt SJ, Quinlan DM. Gender-linked vulnerabilities

to depressive symptoms, stress and problem behaviors in adolescents.

J Res Adolesc 1995;5:1–29.

[20] Huisman M, Oldehinkel AJ, de Winter A, et al. Cohort profile: The

Dutch "Tracking Adolescents’ Individual Lives’ Survey"; TRAILS.

Int J Epidemiol 2008;37:1227–35.

[21] Achenbach TM. Manual for the Child Behavior Checklist/4-18 and

1991 Profile. Burlington, VT: University of Vermont, 1991.

[22] Achenbach TM. Manual for the Youth Self-Report and 1991 Profile.

Burlington, VT: University of Vermont, 1991.

[23] Achenbach TM. Manual for the Teachers’ Report Form and 1991

Profile. Burlington, VT: University of Vermont, 1991.

[24] Verhulst FC, Van der Ende J, Koot HK. Handleiding voor de Youth

Self-Report (YSR). Rotterdam, The Netherlands: Sophia Children’s

Hospital Rotterdam, 1997.

[25] Verhulst FC, Van der Ende J, Koot HK. Handleiding voor de CBCL/

4-18. Rotterdam, The Netherlands: Sophia Children’s Hospital

Rotterdam, 1996.

[26] Noordhof A, Oldehinkel AJ, Verhulst FC, Ormel J. Optimal use of

multi-informant data on co-occurrence of internalizing and external-

izing problems. Int J Methods Psychiatr Res 2008;17:174–83.

[27] Verhulst FC, Koot HM, Van der Ende J. Differential predictive value of

parents’ and teachers’ reports of childrens’ problem behaviors: A longi-

tudinal study. J Abnorm Child Psychol 1994;22:531–46.

[28] Olweus D. Bullying at School. Malden, MA: Blackwell Publishers,

1993. Inc.

[29] Hammen C. Stress generation in depression: Reflections on origins,

research, and future directions. J Clin Psychol 2006;62:1065–82.

[30] Aiken LS, West SG. Multiple Regression: Testing and Interpreting

Interactions. Newbury Park, CA: Sage, 1991.

[31] Gershon A, Minor K, Hayward C. Gender, victimization, and psychi-

atric outcomes. Psychol Med 2008;38:1377–91.

[32] Rudolph KD, Hammen C. Age and gender as determinants of stress

exposure, generation and reactions in youngsters: A transactional

perspective. Child Dev 1999;70:660–77.

[33] Wagner BM, Compas BE. Gender, instrumentality, and expressivity:

Moderators of the relation between stress and psychological symptoms

during adolescence. Am J Community Psychol 1990;18:383–406.

[34] Nadeem E, Graham S. Early puberty, peer victimization and internal-

izing symptoms in ethnic minority adolescents. J Early Adolesc

2005;25:197–222.

[35] Waylen A, Wolke D. Sex ‘n’ drugs ‘n’ rock ‘n’ roll: The meaning and

social consequences of pubertal timing. Eur J Endocrinol 2004;151:

U151–9.

[36] Dohrenwend BP. Inventorying stressful life events as risk factors for

psychopathology: Toward resolution of the problem of intracategory

variability. Psychol Bull 2006;132:477–95.

[37] Monroe SM. Modern approaches to conceptualizing and measuring

human life stress. Annu Rev Clin Psychol 2008;4:33–52.

[38] Wagner C, Abela JRZ, Brozina K. A comparison of stress measure in

children and adolescents: A self-report checklist versus an objectively

rated interview. J Psychopathol Behav Assess 2006;28:251–61.

[39] Oldehinkel AJ, Hartman CA, De Winter AF, et al. Temperament

profiles associated with internalizing and externalizing problems in

preadolescence. Dev Psychopathol 2004;16:421–40.

[40] Berger C, Rodkin PC. Male and female victims of male bullies: Social

status differences by gender and informant source. Sex Roles 2009;61:

72–84.