Content uploaded by Helinä Hakko
Author content
All content in this area was uploaded by Helinä Hakko
Content may be subject to copyright.
Fax +41 61 306 12 34
E-Mail karger@karger.ch
www.karger.com
Original Paper
Ps yc ho pa th ol og y 20 09;42:131–138
DOI: 10.1159/000204764
Bullying Behavior Is Related to Suicide
Attempts but Not to Self-Mutilation
among Psychiatric Inpatient Adolescents
Anu-Helmi Luukkonen a Pirkko Räsänen a Helinä Hakko b Kaisa Riala b
STUDY-70 Workgroup
a Department of Psychiatry, University of Oulu, and
b Department of Psychiatry, Oulu University Hospital,
Oulu , Finland
sions: Among girls, being bullied or bullying others are both
potential risk factors for suicidal behavior. Psychiatric assess-
ment and treatment should thus be considered not only for
victims of bullying, but also for bullies. Suicide-prevention
programs should also routinely include interventions to re-
duce bullying. However, the generalization of our findings
to all adolescents is limited because our study sample con-
sisted of psychiatric adolescent patients. In addition, some
of the possible findings might have remained statistically in-
significant due to the small sample size among adolescents
who had performed suicide attempts or self-mutilation.
Co pyright © 2009 S. Karg er AG, Basel
Introduction
Bullying is a common problem. In Sweden, 18% of the
adolescents in junior high school report that they bully
others, while 10% say that they are bullied. As many as
9% of students report being both a bully and a victim
[1] .
Bullying is defined as aggressive behavior or intentional
harm-doing, which is performed repeatedly and over
time, and which involves an imbalance in power
[2] . Bul-
lying can be physical, verbal or relational
[3] . Being a vic-
tim of bull ying is know n to be related to an inc rea sed ris k
of suicidality compared to those who are not involved in
Key Words
Bullying ⴢ Suicide ⴢ Self-mutilative behavior ⴢ Nonsuicidal
self-injury
Abstract
Background: To investigate the association of bullying be-
havior with suicide attempts and self-mutilation among ad-
olescents. Sampling and Methods: The study sample con-
sisted of 508 Finnish adolescents (age 12–17 years) admitted
to psychiatric inpatient care between April 2001 and March
2006. DSM-IV psychiatric diagnoses and variables measuring
suicidal behavior (i.e. suicide attempts and self-mutilation)
and bullying behavior (i.e. a victim, a bully or a bully-victim)
were obtained from the Schedule for Affective Disorder and
Schizophrenia for School-Age Children Present and Lifetime
(K-SADS-PL). Logistic regression analyses were conducted to
examine the impact of being a victim, a bully or both a bully
and a victim on suicide attempts and self-mutilation. Re-
sults: After adjusting for age, school factors, family factors
and psychiatric disorders, there was a higher risk of suicide
attempts in girls who were victims of bullying (OR = 2.07,
CI = 1.04–4.11, p = 0.037) or who bullied others (OR = 3.27,
CI = 1.08–9.95, p = 0.037). Corresponding associations were
not found for boys; nor was any association of bullying be-
havior with self-mutilation found among either sex. Conclu-
R eceived: October 12, 2007
Acce pted after revision: Ju ne 5, 2008
Pu blished online: Febru ary 27, 2009
Anu-Helmi Luukkonen
University of Oulu, Depart ment of Psychiatry
PO B ox 5000
FI–90014 University of Oulu (Finland)
Tel. +358 8 315 2011, Fax +358 8 333 167, E-Mai l anuhelmi@paju.oulu.f i
© 2009 S. Karger AG, Basel
0254–4962/09/0422–0131$26.00/0
Accessible online at:
www.karger.com/psp
Luukkonen /Räsänen /Hakko /Riala /
STUDY-70 Workgroup
Psychopathology 2009;42:131–138
132
bullying behavior [4, 5] . Recent studies have shown that
bullies are also at a greater risk of suicidal behavior
[4, 5] ,
and that the most troubled adolescents who are at the
greatest risk of suicidality are bully-victims (i.e. those
who are both a bully and a victim of bullying) [4–6] .
Gender differences in both bullying behavior and its
relation with suicidality are significant. Bullying behav-
ior in any form is more prevalent among boys than girls
(47 vs. 30%)
[5] . Boy s als o te nd to b e more common ly bul-
lies and girls more commonly victims
[1, 5] . Interestingly,
girls seem to be more sensitive to bullying than boys
[4–
6] . Klomek et al. [5] found that there is a stronger asso-
ciation between bullying and suicidality among girls
compared to boys.
Earlier studies clearly indicate that bullying behavior
is associated with suicide ideation
[4, 6] and suicide at-
tempts
[5] . However, to our knowledge, none of the previ-
ous studies have examined the association between bul-
lying behavior and self-mutilative behavior. This is sur-
prising,
as the etiology and the intentions behind suicide
attempts and self-mutilation are significantly different
[7] . The intent of self-injurious behavior (often used syn-
onymously with self-mutilation) is typically to avoid or
remove distress and to feel better, but not to eliminate life,
as in suicide attempts. Self-injurious behavior is also
more repetitive in nature and has lower lethality than sui-
cide attempts. However, the terminology and definitions
of this behavior are not consistent
[8] . According to Co-
lumbian Classification Algorithm of Suicide Assessment,
self-injurious behavior is classified as nonsuicidal behav-
ior with nonsuicidal or unknown suicidal intent
[9] .
Keeping in mind this complexity of intentions behind
self-mutilation and its difference compared to suicide at-
tempts, it is clear that both suicide attempts and self-mu-
tilation need to be investigated in relation to bullying be-
havior. We were able to approach these issues by using a
population-based clinical sample of underage adoles-
cents from the geographically large area of northern Fin-
land.
M e t h o d s
Study Population
This study is part of the STUDY-70 project, with data collect-
ed between April 2001 and March 2006. The study sample con-
sisted of 508 adolescents (300 girls, 208 boys) aged 12–17 years
admitted to Unit 70 at Oulu University Hospital Department of
Psychiatry. The majority (98.4%) of the adolescents in the study
population were Caucasians, 1.6% were of other origin. The study
project has been described earlier in detail
[10] . The catchment
area of Unit 70 covers the geographically large area of northern
Finland (the provinces of Oulu and Lapland). All underage ado-
lescents from this area i n need of acute psychiatric hospitaliz ation
are initially treated in Unit 70. The study protocol was approved
by the Ethics Committee of the University of Oulu, Finland.
I n s t r u m e n t s
The subjects were interviewed to obtain DSM-IV diagnoses
using the semi-structured Schedule for Affective Disorder and
Schizophrenia for School-Age Children Present and Lifetime (K-
SADS-PL). According to several studies, K-SADS-PL is known to
be a reliable method for defining DSM-IV diagnoses
[11–13] . The
interview was complemented by interview ing the parents in cases
where data were missing or remained unreliable after interview-
ing the patient.
All data for the purpose of this study, except for some of the
most recent grade point average values at school, were collected
from K-SADS-PL. The grade point averages that were missing
from K-SADS-PL were obtained using the European Addiction
Severity Index (EuropASI), performed by the staff of Unit 70. Eu-
ropASI is an objective fa ce-to-face interview conta ining questions
on general in formation, med ical status, a lcohol and drug use, em-
ployment/support, family and social relationships, legal status
and psychiatric status
[14] .
Suicide Attempts and Self-Mutilation
The information on suicide attempts and self-mutilative be-
havior of each adolescent was obtained using K-SADS-PL. Crite-
ria for a suicide attempt were fulfilled if an adolescent had per-
formed at least 1 suicide attempt with definite suicidal intent or if
it was life-threatening (for example, the adolescent was brief ly
unconscious). Adolescents were defined as having engaged in
self-mutilative behavior if nonsuicidal physical self-damaging
acts had occurred at least 4 times during the last year, or if they
had caused serious injury (e.g. broken bone or burn with scar-
ring). In this study, suicidal behavior was analyzed in 3 exclusive
subgroups: (1) nonsuicidal behavior, (2) subjects with self-mutila-
tion, but without suicide attempts and (3) subjects with suicide
attempts. Subjects with both self-mutilation and suicide attempts
(n = 49, 9.6%) were included in the suicide attempts group.
B u l l y i n g
The data on bullying behavior were gathered from 2 sections
of the K-SADS-PL. Whether an adolescent was bullied or not was
asked in the nonstructured part (school adaptation and social re-
lations) of the K-SADS-PL. Bullying others was gathered from K-
SADS-PL criter ia for conduct dis order, and was defi ned as present
if an adolescent had threatened or intimidated another on 3 or
more occasions. In this study, bullying behavior was categorized
into the following 4 subgroups: (1) bullies, (2) victims, (3) both
bully and victims ( bully-victims) and (4) adolescents not involved
in bullying behavior.
School Factors
School factors were obtained in the nonstructured part of the
K-SADS-PL. Adolescents were asked whether they had ever re-
peated grade(s), used special services or had teachers complain
about their behavior. The most recent grade point average was
also enquired about in this part of the interview. If this informa-
tion was not available from the K-SADS-PL, it was complemented
by information from the EuropASI.
Bullying and Suicidal Behavior Psychopathology 2009;42:131–138
133
Family Factors
K-SADS-PL anamnesis was used to define the family type of
an adolescent (2 biological parents, 1 biological parent, other). It
was also used to determine whether domestic violence, physical
or sexual abuse had occurred (yes, no). Domestic violence was
recorded if the adolescent had ever witnessed explosive argu-
ments at home, involving threatened or actual harm to parent.
Physical abuse was defined as present when the adolescent had
been injured (bruises on more than 1 occasion or more serious
injury) as a resu lt of physical assault by a parent. Sexual abuse was
defined as present if the adolescent had ever experienced unwant-
ed isolated or repeated incidents of genital fondling, oral sex, or
vaginal or anal intercourse.
Psychiatric Disorders
In this study, we used DSM-IV-based psychiatric d iagnoses of
the adolescents that were obtained by using K-SADS-PL. Six ma-
jor diagnostic categories for current psychiatric disorders were
used in this study: substance-related disorders (DSM-IV: 303.9,
304.0–304.6, 304.8–304.9, 305.0, 305.2–305.7, 305.9), affective
disorders (296.2–296.3, 300.4, 311), anxiety disorders (300.00–
300.02, 300.21–300.23, 300.29, 300.3, 308.3, 309.81), conduct dis-
order s (312.8–312.9, 313.81, 314.00–314.01, 314.9, 299.8), ps ychot-
ic disorders (295, 296.0, 296.4–296.9, 297.1–297.3, 298.8–298.9,
301.13, 301.22) and other psychiatric disorders than those men-
tioned above. These diagnoses were overlapping, i.e. a patient
might have various types of psychiatric diagnoses.
S t a t i s t i c a l M e t h o d s
Statistical significance of group differences was assessed with
Pearson’s 2 test. The association of bullying behavior with sui-
cidal behavior in adolescents was examined with a logistic regres-
sion model after adjusting for age, school factors, family factors
and psychiatric disorders. The statistical software used was SPSS
for Windows version 14.0.
R e s u l t s
Gender Differences in Bullying Behavior and Suicidal
Behavior
Of all adolescent boys, 91 (43.8%) had no bullying be-
havior, while 61 (29.3%) were victims of bullying, 42
(20.2%) had bullied others and 14 (6.7%) were bully-vic-
tims. In adolescent girls, these proportions were 141
(47.0%), 115 (38.3%), 30 (10.0%) and 14 (4.7%), respective-
ly. There was a statistically significant difference in bul-
lying behavior between genders ( 2 = 13.1, d.f. = 3, p =
0.004). Regarding suicidal behavior, of all adolescent
boys, 161 (77.4%) were nonsuicidal, 21 (10.1%) had per-
formed self-mutilation and 26 (12.5%) had attempted sui-
cide, while the respective proportions in adolescent girls
were 148 (49.3%), 74 (24.7%) and 78 (26.0%) ( 2 = 40.8,
d.f. = 2, p ! 0.001).
Significant Gender Differences in School and Family
Factors and Psychiatric Disorders
In our study, approximately every fifth (20.7%) boy
had repeated a grade(s), whereas less than every tenth girl
(8.7%) had repeated a grade(s) ( 2 = 15.1, d.f. = 1, p !
0.001). Special services were also more often used by boys
(62.5%) than girls (42.3%) ( 2 = 20.0, d.f. = 1, p ! 0.001).
Of the girls 44.0% and of the boys 27.9% were living
in a family with 2 biological parents ( 2 = 13.9, d.f. = 2,
p = 0.001). Sexual abuse was reported over 6 times more
often by girls (22.7%) compared to boys (3.4%) ( 2 = 36.4,
d.f. = 1, p ! 0.001).
Of the psychiatric disorders used in this study, gender
differences were seen in the prevalence of affective disor-
ders, anxiety disorders, conduct disorders and other dis-
orders. Thus, 55.3% of the girls had affective disorder,
whereas its prevalence among boys was only 35.1% ( 2 =
20.2, d.f. = 1, p ! 0.001). Anxiety disorders were also more
common among girls (29.0%) than boys (16.3%) ( 2 =
10.8, d.f. = 1, p = 0.001). In contrast, the prevalence of
conduct disorders was higher among boys (58.7%) than
among girls (35.0%) ( 2 = 27.8, d.f. = 1, p ! 0.001). Other
disorders, including eating disorders, were over twice
as common among girls (18.3%) than among boys (7.2%)
( 2 = 12.8, d.f. = 1, p ! 0.001).
Background Variables in Relation to Suicidal
Behavior of Girls and Boys
Tables 1 and 2 show the bivariate associations be-
tween background variables and suicidal behavior for
boys and girls, respectively. Among girls, affective dis-
orders ( 2 = 36.9, d.f. = 2, p ! 0.001), anxiety disorders
( 2 = 17.7, d.f. = 2, p ! 0.001) and sexual abuse ( 2 = 7.9,
d.f. = 2, p = 0.019) were statistically significantly associ-
ated with suicidal behavior. Among boys, the only sig-
nificant association was found between suicidal behav-
ior and affective disorders ( 2 = 18.9, d.f. = 2, p !
0.001).
Bullying Behavior and Psychiatric Disorders
Bullying behavior was significantly associated with
substance-related disorder and conduct disorder among
both sexes. Among girls ( 2 = 27.2, d.f. = 3, p ! 0.001) and
boys ( 2 = 14.8, d.f. = 3, p = 0.002), the victims were sta-
tistically significantly less often suffering from sub-
stance-related disorder than those not involved in bully-
ing behavior. The prevalence of conduct disorder was
highest among bully-victims among both girls ( 2 = 55.1,
d.f. = 3, p ! 0.001) and boys ( 2 = 42.0, d.f. = 3, p !
0.001).
Luukkonen /Räsänen /Hakko /Riala /
STUDY-70 Workgroup
Psychopathology 2009;42:131–138
134
Association of Bullying Behavior with Self-Mutilation
and Suicide Attempts
Table 3 presents the results of a logistic regression
analysis in which the association of bullying behavior
with self-mutilation and suicide attempts was investigat-
ed after adjusting for age, school factors, family factors
and psychiatric disorders (see confounders in tables 1 and
2 ). As seen in table 3 , being a victim was statistically sig-
nificantly associated with suicide attempts among girls
(OR = 2.07, CI = 1.04–4.11, p = 0.037). Furthermore, girls
who bullied others had over a 3-fold risk of suicide at-
tempts (OR = 3.27, CI = 1.08–9.95, p = 0.037) compared
with those not involved in bullying behavior. No associa-
tio n of bu ll yin g b ehav ior w it h s ui cide at tempts wa s fou nd
among boys. An association between being a bul ly-victim
and suicidal behavior was not found among either boys
or girls, nor was any association of bullying behavior with
self-mutilation found among either sex.
Of all covariates used in logistic regression analyses,
affective disorders were associated with self-mutilation
both among boys (OR = 4.40, CI = 1.68–11.55, p = 0.003)
and girls (OR = 3.85, CI = 2.09–7.07, p ! 0.001). Affective
disorders were also associated with suicide attempts both
among boys (OR = 4.67, CI = 1.82–12.03, p = 0.001) and
Tab le 1. Background variables of male adolescent psychiatric inpatients according to subgroups of suicidal be-
havior
Boys suicide groups p value1
non-suicidal
(n = 161)
self-mutilation
(n = 21)
suicide attempts
(n = 26)
n% n% n%
Bullying factors
No bullying behavior 74 46.0 10 47.6 7 26.9 0.647
Victim 45 28.0 5 23.8 11 42.3
Bully 31 19.3 5 23.8 6 23.1
Both bully and victim 11 6.8 1 4.8 2 7.7
School factors2
Repeated grade(s) 31 19.3 6 28.6 6 23.1 0.580
Special services 102 63.4 14 66.7 14 53.8 0.596
Teacher’s complaint about behavior 83 51.6 9 42.9 13 50.0 0.754
Family factors
Family type
2 biological parents 49 30.4 2 9.5 7 26.9 0.283
1 biological parent 55 34.2 9 42.9 7 26.9
Other 57 35.4 10 47.6 12 46.2
Domestic violence244 27.3 6 28.6 9 34.6 0.746
Physical abuse239 24.2 9 42.9 4 15.4 0.086
Sexual abuse26 3.7 0 0 1 3.8 0.665
Psychiatric disorders2
Substance-related disorder 66 41.0 10 47.6 10 38.5 0.803
Affective disorder 44 27.3 13 61.9 16 61.5 0.000
Anxiety disorder 25 15.5 3 14.3 6 23.1 0.605
Conduct disorder 95 59.0 14 66.7 13 50.0 0.505
Psychotic disorder 22 13.7 2 9.5 3 11.5 0.845
Other 11 6.8 2 9.5 2 7.7 0.900
Mean (std) grade point average 6.9 0.8 7.0 1.0 7.0 0.7 0.6443
Mean (std) age at admission, years 15.4 1.4 15.4 1.5 15.5 1.4 0.8543
1 Pearson’s 2 test, two-tailed significance.
2 Positive responses.
3 One-way ANOVA.
Bullying and Suicidal Behavior Psychopathology 2009;42:131–138
135
girls (OR = 6.34, CI = 3.20–12.58, p ! 0.001). Among girls
sexual abuse (OR = 2.69, CI = 1.27–5.69, p = 0.010) and
anxiety disorder (OR = 2.61, CI = 1.35–5.04, p = 0.004)
also increased the likelihood of suicide attempts.
Discussion
To the authors’ knowledge, this is the first study so far
to investigate the impact of bullying behavior on both
self-mutilation and suicide attempts. Our major finding
was that, among girls, both being a victim and bullying
ot he rs i nc rea se d t he r is k of sui cide at tempts ove r t wof old .
This finding is consistent with previous studies
[4– 6] ,
which state that girls are more sensitive to bullying than
boys. Contrary to our expectations, bullying behavior
was not associated with self-mutilation among either sex.
For girls, our major finding replicates the earlier reports
that bullies, not just victims, are at higher risk of suicidal
behavior
[4, 5] . In the study of Klomek et al. [5] , an asso-
ciation was found between being bullied or bullying oth-
ers and suicide attempts (OR 3.6–4.5) among both girls
and boys. In their study, the association between bullying
behavior and suicide attempts was stronger than in our
Tab le 2. Background variables of female adolescent psychiatric inpatients according to subgroups of suicidal
behavior
Girls suicide groups p value1
non-suicidal
(n = 148)
self-mutilation
(n = 74)
suicide attempts
(n = 78)
n% n% n%
Bullying factors
No bullying behavior 82 55.4 32 43.2 27 34.6 0.145
Victim 48 32.4 30 40.5 37 47.4
Bully 12 8.1 8 10.8 10 12.8
Both bully and victim 6 4.1 4 5.4 4 5.1
School factors2
Repeated grades 12 8.1 7 9.5 7 9.0 0.939
Special services 58 39.2 32 43.2 37 47.4 0.483
Teacher’s complaint about behavior 62 41.9 34 45.9 31 39.7 0.733
Family factors
Family types
2 biological parents 74 50.0 32 43.2 26 33.3 0.091
1 biological parent 41 27.7 21 28.4 22 28.2
Other 33 22.3 21 28.4 30 38.5
Domestic violence244 29.7 21 28.4 27 34.6 0.665
Physical abuse241 27.7 17 23.0 21 26.9 0.745
Sexual abuse225 16.9 17 23.0 26 33.3 0.019
Psychiatric disorders2
Substance-related disorder 50 33.8 26 35.1 32 41.0 0.550
Affective disorder 56 37.8 51 68.9 59 75.6 0.000
Anxiety disorder 35 23.6 15 20.3 37 47.4 0.000
Conduct disorder 55 37.2 31 41.9 19 24.4 0.057
Psychotic disorder 16 10.8 8 10.8 3 3.8 0.181
Other 27 18.2 15 20.3 13 16.7 0.847
Mean (std) grade point average 7.6 0.9 7.4 0.9 7.5 1.0 0.3213
Mean (std) age at admission, years 15.4 1.4 15.3 1.1 15.8 1.2 0.0223
1 Pearson’s 2 test, two-tailed significance.
2 Positive responses.
3 One-way ANOVA.
Luukkonen /Räsänen /Hakko /Riala /
STUDY-70 Workgroup
Psychopathology 2009;42:131–138
136
study, even though their study population consisted of
school pupils. This difference could be explained by the
different criteria used in the studies to define a suicide
attempt: in the study of Klomek et al.
[5] , a suicide at-
tempt was defined as present if the adolescent reported
any past attempt, regardless of injury or need for medical
attention. In the population-based study of Kaltiala-Hei-
no et al.
[4] , it was found that among girls both being bul-
lied and being a bully was associated with suicidal ide-
ation, while among boys an association was only found
between being a victim of bullying and suicidal ide-
ation.
No association of bullying behavior with suicide at-
tempts or self-mutilation was found among boys in our
study. In previous studies
[5, 6] , an association between
bullying behavior and suicidal behavior has been found
among boys, but also in these studies the association was
weaker among boys than girls. We think that there are
several reasons why an association of bullying behavior
with suicidal behavior can be seen among girls, but not
among boys. Firstly, in our opinion, boys who are bullied
do not necessarily report being bullied as easily as girls,
because boys might be more ashamed of being victims
than girls. Secondly, we support the idea of Klomek et al.
[5] , who have explained this gender difference by a gender
paradox in which girls are more rarely bullies, but, when
they are, they have more severe impairment than boys.
This same phenomenon is seen in the findings of Was-
serman et al.
[15] stating that girls who have conduct
problems are at a higher risk of affective disorders than
boys. Thirdly, the number of boys (208) in our sample is
much smaller than the number of girls (300). This might
be due to the fact that even though suicide attempts in
Europe are more common among females than among
males
[16] , the suicide mortality rate is higher among
males than females [17, 18] . In Finland, suicide as a cause
of death is over 3 times more common among males than
among females (31.7 per 100,000 mean population vs. 9.4
per 100,000 mean population)
[18] . Thus, the small num-
ber of boys in our sample might be explained by the fact
that more boys than girls succeed in their suicide at-
tempts and these boys are thus not seen in our study sam-
ple.
Possible confounders of the relationships found in our
study were behavioral problems, which cannot be as-
sessed by the K-SADS-PL. We were not able to assess per-
sonality features such as impulsivity, aggressiveness and
identity confusion, which are typical for developing per-
sonality disorders. Additionally, it is a clinical practice in
Finland that borderline personality disorder (BPD) and
other personality disorders are not usually diagnosed un-
der the age of 18. However, BPD might act as mediating
factor between bullying and suicidal behavior in girls, as
BPD, which is more common among females than among
males in clinical samples
[19, 20] , has been shown to be
related with suicidal behavior
[21, 22] .
Limitations and Strengths
The main limitation of the present study is the rather
small number of adolescents who had performed suicide
at te mpt s or s elf-mut il ation. In ou r stu dy, so me o f t he p os-
sible findings might have remained statistically insignif-
icant due to the small sample size (type II error). Thus,
there is a need to study further the association of being a
victim of bullying with self-mutilation, for example. Es-
pecially the findings among boys and bully-victims
should be interpreted with caution due to the small num-
ber of suicide attempts and self-mutilation among boys
and the small number of bully-victims
in our sample.
Fu rt herm ore , we were not ab le t o stu dy b ul ly ing beh avior
among subjects with both self-mutilation and suicide at-
tempts due to the small sample size.
The generalization of our findings to all adolescents is
limited because our study sample consisted of psychiatric
Tab le 3. The association of bullying behavior to self-mutilation
and suicide attempts of male and female adolescent psychiatric
inpatients
Self-mutilation vs.
non-suicidal
Suicide attempts vs.
non-suicidal
OR and 95% CI p
value
OR and 95% CI p
value
Boys
No bullying behavior reference reference
Bully 1.25 (0.38–4.11) 0.717 2.16 (0.59–7.87) 0.244
Victim 0.83 (0.26–2.67) 0.753 2.72 (0.88–8.41) 0.081
Both bully and victim 1.07 (0.12–9.87) 0.953 3.14 (0.51–19.50) 0.220
Girls
No bullying behavior reference reference
Bully 2.10 (0.75–5.91) 0.160 3.27 (1.08–9.95) 0.037
Victim 1.70 (0.89–3.26) 0.107 2.07 (1.04–4.11) 0.037
Both bully and victim 1.56 (0.39–6.30) 0.532 1.93 (0.40–9.37) 0.415
A logistic regression analysis predicting suicidal behavior with bully-
ing behav ior (method = enter) af ter adjusting for age, school factors (g rade
point average, repeated grades, special services, teacher’s complaint about
behavior), family factors (family type, domestic violence, physical abuse,
sexua l abuse), psychiatric disorders (affective disorder, anxiety disorder,
conduct disorder, psychotic disorder and other) (method = forward step-
wise).
Bullying and Suicidal Behavior Psychopathology 2009;42:131–138
137
adolescent patients. On the other hand, our study sample
comprises all adolescent patients hospitalized in the geo-
graphically large area of northern Finland, and thus rep-
resents the most serious cases in the general adolescent
population.
The age at which the adolescents have been involved
in bullying behavior remains unknown, as the age of the
adolescent is only known at the time of the interview.
This might affect the results, as it has been suggested that
those adolescents who are still involved in bullying be-
havior at an older age are more disturbed than those who
engage in bullying behavior when younger and when bul-
lying is more normative
[5] . In addition, the questions
used in the interview do not allow us to properly estimate
the severity of the bullying behavior.
The strength of this study was that all the data were
gathered using established semi-structured interviews
[12, 14] .
In summary, our study clearly shows that among girls
both being a victim of bullying and bullying others are
significant risk factors for suicidal behavior in adoles-
cents, independent of other risk factors. Both teachers
and school nurses should pay more attention to pupils
who are involved in bullying behavior. Mental well-being
of both victims of bullying and bul lies should be screened
in primary health care. Subjects with suicidality should
be further directed to specialist-level child and adoles-
cent psychiatric evaluation. Suicide prevention programs
should also routinely include interventions to reduce bul-
lying.
A c k n o w l e d g m e n t s
This study received support from the Päivikki and Sakari
Sohlberg Foundation, Ethel F. Donaghue Women’s Health Inves-
tigator Program at Yale, Alcoholic Beverage Medical Research
Foundation (ABMRF), VA MIRECC, the Alma and K.A. Snell-
man Foundation and Yrjö Jahnsson Foundation.
We also thank the staff of the adolescent Unit 70 of the Oulu
University Hospital for data collection, and all the adolescents
who participated in this study.
A p p e n d i x
STUDY-70 Workgroup
Medical supervisors: Prof. Pirkko Räsänen, University of
Oulu, Department of Psychiatry; Prof. Jaakko Lappalainen, Yale
University, School of Medicine, Department of Psychiatry; Prof.
Mauri Marttunen, Helsinki University Central Hospital, Depart-
ment of Adolescent Psychiatry, Peijas. Administrative supervi-
sors: Kaisa Riala, MD, PhD; Taru Ollinen, MD, PhD; Kristian
Läksy, MD, PhD, Oulu University Hospital, Department of Psy-
chiatry. Statistical consultants: Helinä Hakko, PhD; Kaisa Kar-
vonen, BSc, Oulu University Hospita l, Depart ment of Psychiatry.
Senior researchers: Prof. Markku Timonen, University of Oulu,
Department of Public Health Sciences and General Practice. Re-
searchers: Essi Ilomäki, MD; Risto Ilomäki, MD; Vesa Tikkanen,
medical student, University of Oulu, Department of Psychiatry;
Matti Lauk kanen, MSc, Oulu University Hospital, Department of
Psychiatry.
References
1 Ivarsson T, Broberg AG, Arvidsson T, Gill-
berg C: Bullying in adolescence: psychiatric
problems in vic tims and bull ies as measured
by the Youth Self Report (YSR) and the De-
pression Self-Rating Sca le (DSRS). Nord J
Psychiatry 2005;
59: 365–373.
2 Olweus D: Sweeden; in Smith PK, Morita Y,
Junger-Tas J, Olweus D, Catalona R, Slee P
(eds): The Nature of School Bullying: A
Cross-National Perspective. New York,
Routledge, 1999, pp 7–27.
3 Fekkes M, Pijpers FI, Verloove-Vanhorick
SP: Bullying: who does what, when and
where? Involvement of children, teachers
and parents in bullying behavior. Health
Educ Res 2005;
20: 81–91.
4 Kalt iala-Heino R, R impela M, Mart tunen M,
Rimpelä A, Rantanen P: Bullying, depres-
sion, and suicidal ideation in Finnish adoles-
cents: school survey. BMJ 1999;
319: 348–351.
5 Klomek AB, Marrocco F, Kleinman M,
Schonfeld IS , Gould M: Bul lying, depres sion,
and suicidality in adolescents. J Am Acad
Child Adolesc Psychiatry 2007;
46: 40–49.
6 Kim YS, Koh YJ, Leventhal B: School bully-
ing and su icidal risk i n Korean middle sc hool
students. Pediatrics 2005;
115: 357–363.
7 Skegg K: Self-harm. Lancet 2005;
366: 1471–
148 3.
8 Muehlenkamp JJ: Self-injurious behavior as
a separate clinical syndrome. Am J Ortho-
psychiatry 2005;
75: 324–333.
9 Posner K, Oquendo MA, Gould M, Stanley
B, Davies M: Columbia Classification Algo-
rithm of Suicide Assessment (C-CASA):
classification of suicidal events in the FDA’s
pediat ric suicidal r isk analysis of a ntidepres-
sants. A m J Psychiatr y 2007;
164: 1035–1043.
10 Mä kikyrö T, Hakko H, Timonen MJ, Lappa-
lainen JAS, Ilomäki RS, Marttunen MJ,
Läk sy K, Räsänen PK: Smok ing and suicid al-
ity among adolescent psychiatric patients. J
Adolesc Health 2004;
34; 250–253.
11 Ambrosini PJ: Historical development and
present status of the Schedule for Affective
Disorders a nd Schizophren ia for School-Age
Children (K-SADS). J Am Acad Child Ado-
lesc Psychiatry 2000;
39:49–58.
12 Kaufman J, Birmaher B, Brent D, Rao U,
Flynn C, Moreci P, Williamson D, Ryan N:
Schedule for A ffecti ve Disorders and Sch izo-
phrenia for School-Age Children – Present
and Lifetime Version (K-SADS-PL): initial
reliabi lity and va lidity dat a. J Am Acad Chi ld
Adolesc Psychiatry 1997;
36: 980–988.
Luukkonen /Räsänen /Hakko /Riala /
STUDY-70 Workgroup
Psychopathology 2009;42:131–138
138
13 K im YS, Cheon KA , Kim BN, Chang SA, Yoo
HJ, Kim J W, Cho SC, Seo DH, Bae MO, So
YK, Noh JS, Koh Y J, McBurnett K, L eventhal
B: The reliability and validity of Kiddie-
Schedule for A ffecti ve Disorders and Sch izo-
phrenia – Pre sent and Lifet ime Version – Ko-
rean Version (K-SADS-PL-K). Yonsei Med J
2004;
29: 81–89.
14 Kokkevi A, Hartgers C: European adapta-
tion of a multidimensional assessment in-
strument for drug and alcohol dependents.
Euro Addict Res 1995;
1:208–210.
15 Wass erman GA, McRey nolds LS, Ko SJ, Katz
LM, Car penter JR: Gender differences in
psychiatric disorders at juvenile probation
intake. Am J Public Health 2005;
95: 131–
137.
16 Bille-Brahe U, Kerkhof A, De Leo D,
Schmidt ke A, Crepet P, Lonnqvist J, Michel
K, Salander-Renberg E, Stiles TC, Wasser-
mann D, Aagaard B, Egebo H, Jensen B: A
repetition-prediction study of European
parasuicide populations: a summar y of the
first report from par t II of the WHO/EURO
Multicentre Study on Parasuicide in co-op-
eration with the EC concerted action on at-
tempted suicide. Acta Psychiatr Scand 1997;
95:81–86.
17
Suicide deat h rate, by age group, males a nd fe-
males 2006. http://epp.eurostat.ec.europa.eu
18 Statistics Finland: Statistical Yearbook of
Finland 2006. Hämeenlinna, Karisto Oy,
2006, vol 101, p 529.
19 Gunderson JG: Borderline Personality Dis-
order: A Cli nical Guide. Washi ngton, Amer-
ican Psychiatric Press, 2001.
20 Kantojär vi L, Veijola J, Läksy K, Jokelainen
J, Herva A, Karvonen JT, Kokkonen P, Jär-
velin MR, Joukama a M: Comparison of hos-
pital-treated personality disorders and per-
sonality disorders in a general population
sample. Nord J Psychiatry 2004;
58: 357–362.
21 Oldham J M: Borderline pers onality di sorder
and suicidality. Am J Psychiatry 20 06;
163:
20–26.
22 Oquendo MA, Bongiovi-Garcia ME, Gal-
falvy H, Goldberg PH, Grunebaum MF,
Burke AK, Mann IJ. Sex differences in clini-
cal predictors of suicida l acts after major de-
pression: a prospective study. Am J Psychia-
try 2007;
164: 134–141.
