Journal of Adolescent Health 44 (2009) 464–467

Original article

Mental Disorders in Self-Cutting Adolescents

Jukka Hintikka, M.D.a,b,e,*, Tommi Tolmunen, M.D.c, Marja-Liisa Rissanen, M.H.Scic,
Kirsi Honkalampi, Ph.D.c, Jari Kylmä, Ph.D.d, and Eila Laukkanen, M.D.c,e
a
Department of Psychiatry, Paijat-Hame Central Hospital, Lahti, Finland
b
Department of Psychiatry, Medical School, University of Tampere, Tampere, Finland
c
Department of Psychiatry, Kuopio University Hospital, Kuopio, Finland
d
Department of Nursing Science, University of Kuopio, Kuopio, Finland
e
Department of Psychiatry, Institute of Clinical Medicine, University of Kuopio, Kuopio, Finland
Manuscript received March 12, 2008; manuscript accepted October 8, 2008

Abstract Purpose: Self-cutting as a form of self-harm is common in general population adolescents. The aim
of this study was to investigate the prevalence of mental disorders and associated factors among self-
cutting community-dwelling adolescents.
Methods: A sample of adolescents who reported current self-cutting (n ¼ 80) was drawn from a large
sample of community adolescents (n ¼ 4205). Of these 80 individuals, 44 consented to further detailed
assessment. An age- and gender-matched control subject was selected for each study subject. Data
collection included Structured Clinical Interviews for DSM-IV-TR, the Beck Depression Inventory,
the Alcohol Use Disorders Identification Test (AUDIT), and the Youth Self-Report for adolescents
aged 11–18 years (YSR).
Results: Major depressive disorder (63% vs. 5%), anxiety disorders (37% vs. 12%), and eating
disorders (15% vs. 0%) were more common among self-cutting girls (n ¼ 41) than among controls.
None of the self-cutting boys (n ¼ 3) had a DSM-IV-TR Axis I mental disorder. In multivariate model,
the presence of major depressive disorder, the AUDIT score and the YSR internalizing subscale score
were the factors that were independently associated with the presence of self-cutting in girls.
Conclusions: Major depressive disorder, signs of alcohol misuse, and internalizing behavior strongly
associate with self-cutting in community-dwelling adolescents, especially in girls. Ó 2009 Society for
Adolescent Medicine. All rights reserved.
Keywords: Self-destructive behavior; Self-cutting; Major depressive disorder; Alcohol abuse; Adolescence

Deliberate self-harm has become more prevalent in recent and this has been suggested as one explanation for gender
years among adolescents [1]. Among psychiatric populations, differences in the frequency of and motivations for self-harm-
it is associated with various mental disorders, e.g., depres- ing behaviors [14,15].
sion, bipolar disorder, and borderline personality disorder A recent meta-analysis of self-reported suicidal phenom-
[2–7]. Self-cutting as a form of deliberate self-harm is also ena among adolescents, based on 128 population-based
common among nonclinical adolescents [8]. studies and 518,188 study subjects, revealed that 13% of ad-
In addition to mental disorders, depressive symptoms olescents have engaged in deliberate self-harm at some point
[9,10], heavy alcohol use and other substance misuse [16]. Nevertheless, the authors concluded that methodologi-
[11,12], and poor problem-solving skills [13] are associated cal factors, for example the study population and terminology
with deliberate self-harm. Girls are more prone to internalize used, may influence the results. Thus, although Evans et al
than boys, and boys are more prone to externalize than girls; [16] were able to include an exhaustive number of studies
and study subjects in their meta-analysis, there is still
a need for studies that use terminology carefully and consis-
*Address correspondence to: Jukka Hintikka, M.D., Kuopio University
Hospital, Department of Psychiatry, Building 5, 5th floor, P.O. Box 1777, tently and focus on specific topics. For example, the preva-
FI-70211 Kuopio, Finland. lence of mental disorders is not known in community
E-mail address: [email protected] adolescents who engage in self-cutting.
1054-139X/09/$ – see front matter Ó 2009 Society for Adolescent Medicine. All rights reserved.
doi:10.1016/j.jadohealth.2008.10.003
 J. Hintikka et al. / Journal of Adolescent Health 44 (2009) 464–467 465

In this study our aim was to investigate the prevalence of The final study and control groups each comprised 41 girls
mental disorders and associated factors in a sample of and three boys, with a mean age of 15.2 (SD ¼ 1.5) years.
13–18-year-old Finnish community adolescents who Depressive symptoms were examined by using the Beck
reported current self-cutting, and to compare these adoles- Depression Inventory (BDI) scale [18,19]. This yields a total
cents with gender- and age-matched controls who had never score from 0 to 63, and a score of 10 or more indicates at least
self-cut. We used the definition proposed by Favazza [17], mild depression. The Alcohol Use Disorders Identification
according to which self-mutilation is a deliberate act to Test (AUDIT) was used to screen alcohol misuse [20]. It is
destroy one’s own body tissue without a conscious intent to a structured questionnaire originally designed by the World
die. This definition includes self-cutting with any instruments Health Organization to be used with adults in measuring their
but not tattooing or body piercing. use of alcohol, but it has also been used on young people and
found to be sensitive enough to distinguish problematic use
Methods of alcohol [21].
The competence, adaptive functioning and problems of the
Subjects for this study were drawn from a sample of adolescents were assessed by a standardized self-rated ques-
13–18-year-old boys (n ¼ 1953) and girls (n ¼ 2252) who tionnaire, the Youth Self-Report for ages 11–18 years (YSR)
participated in a study on adolescents attending school in [22]. This includes 20 competence items concerning an adoles-
Kuopio, a city in eastern Finland with approximately cent’s participation in various activities, and 112 items that
90,000 inhabitants. Permission for the study was obtained measure eight subscale symptoms: withdrawn, somatic com-
in advance from the headmasters of the schools. The ethical plaints, anxiety and depression, aggressive behavior, rule-
committee of Kuopio University Hospital and the University breaking behaviors, social problems, thought problems, and
of Kuopio granted permission for the study. attention problems. The first three subscales are referred to
The baseline study focused on comprehensive, secondary, as ‘‘internalizing’’ and the next two as ‘‘externalizing.’’ Over-
and secondary modern schools. Two special schools were all behavioral and emotional functioning are measured by the
excluded from the study by recommendation of their head- total problem scale. The total scale score and internalizing and
masters, as the questionnaires were considered too compli- externalizing subscale scores were assessed in this study.
cated for their disabled students. Before participation in the Psychiatric diagnoses were based on clinical assessments
study, written informed consent was requested from the ado- and were verified for all study subjects by two interviewers
lescents and from the parents of those aged less than 15 years. (M.-L.R., T.T.) using the Structured Clinical Interview for
The baseline data were collected by structured self-rating DSM-IV-TR (SCID-I) [23]. Both interviewers had several
questionnaires that the participants completed during class years of clinical experience. In addition, before the study,
periods at school. Half of the upper secondary schools the interviewers had attended a 3-day course in SCID inter-
(n ¼ 8), two of 10 secondary schools, and one of three viewing. The cases for which the diagnosis was uncertain
secondary modern schools wanted to administer the study were discussed in round-table discussions with two other
by themselves. These students represented 33% of the base- experienced clinicians (J.H., E.L.).
line study subjects (n ¼ 7087) and the response rate was SPSS version 14.0 statistical software was used for data
76.8% (c2 ¼ 99.99, df ¼ 1, p< .001 as compared with other analysis. The differences between the groups of self-cutting
schools). In these cases, the researcher informed the teachers participants and others were evaluated with the Pearson
about the study and gave them instructions on how to perform Chi-square test and Fisher’s two-tailed exact test for
it. In the other schools, a researcher administered the test categorical variables and the two-tailed Student’s t-test or
(67% of study subjects; response rate 61.5%). The feasibility Mann-Whitney U-test (depending on the distribution)
of the questionnaires was tested in a pilot study performed in for continuous variables. Logistic regression models were
one class (n ¼ 27).
The questionnaire about self-cutting included a short Table 1
instruction as follows: ’’A person can consistently harm Total numbers of DSM-IV-TR mental disorders in self-cutting girls and
control girlsa
him/herself, for example, by taking an overdose of medicine
or other drugs. Self-cutting is one way to harm oneself. Now Self-cutting girls Control girls
we are asking you, have you ever engaged in any kind of de- (n ¼ 41) (n ¼ 41)
liberate self-harming behavior?’’ The questions concerning Major depressive disorder 27 2
self-cutting were: ‘‘Have you ever cut yourself (1 ¼ yes; Anxiety disorders 20 6
2 ¼ never)?’’ ‘‘Do you currently cut yourself (1 ¼ yes; Eating disorders 6 –
Conduct disorders 4 2
2 ¼ no)?’’ ‘‘When have you last cut yourself?’’ Alcohol abuse 1 1
In this study, those who had reported current self-cutting Cannabis abuse 1 –
in the baseline questionnaire (n ¼ 80) and who gave their Psychotic disorder NOS 1 –
own or their parents’ consent for further studies were enrolled Total 60 11
for a psychiatric interview (n ¼ 44). The controls, who had NOS ¼ not otherwise specified.
a
never self-cut, were matched according to age and gender. Each subject could have more than one diagnosis (range, 0–5).
466 J. Hintikka et al. / Journal of Adolescent Health 44 (2009) 464–467

constructed to explore which factors explained the presence Table 2

of self-cutting. In general, a p-value<.05 was considered Scale scores in self-cutting girls and control girlsa
statistically significant in the analyses. Self-cutting girls Control girls
Measure (n ¼ 41) Mean (SD) (n ¼ 41) Mean (SD)

Results Beck Depression Inventory 18.0 (10.3) 3.7 (3.5)

AUDIT 6.2 (6.2) 1.7 (2.7)
The total numbers of mental disorders in self-cutting girls YSR (total) 72.1 (28.8) 34.5 (16.6)
Internalizing subscale 27.0 (10.0) 10.0 (5.5)
(n ¼ 41) and their age- and gender-matched controls (n ¼ 41)
Externalizing subscale 19.8 (10.0) 10.7 (5.8)
are presented in Table 1. Mental disorders were more preva-
lent among the self-cutting girls than among their age- AUDIT ¼ Alcohol Use Disorders Identification Test; YSR¼Youth Self-
Report for adolescents aged 11–18 years.
matched controls (79% vs 21%, c2 ¼ 23.74, df ¼ 1, p< a
Difference between the groups in each scale score is statistically highly
.001). The difference in psychiatric comorbidity (at least significant (p < .001).
two psychiatric diagnoses) was at the same level (42% vs
7%, c2 ¼ 12.96, df ¼ 1, p< .00 1). None of the self-cutting
disorders may severely disturb development during adoles-
boys (n ¼ 3) or their controls had a DSM-IV-TR mental
cence. The core psychopathology associated with these
disorder.
disorders comprises symptoms of negative affects, e.g.,
Major depressive disorder (63% vs. 5%, c2 ¼ 31.24, p<
depression, anxiety, and anger. Self-cutting may be used to
.001), anxiety disorders (37% vs. 12%, c2 ¼ 6.61, p ¼ .010)
alleviate these affects [24]. None of the self-cutting boys
and eating disorders (15% vs. 0%, Fisher’s exact test,
(n ¼ 3) or their controls had a DSM-IV-TR mental disorder.
p ¼ .026) were more common among self-cutting girls than
However, because of the small number of boys in the sample,
among controls. Moreover, the self-cutting girls had higher
this study does not allow any conclusions to be drawn about
levels of psychopathology than the controls in all scales
associations between self-cutting and mental disorders in
used in assessments (Table 2). In self-cutting and control
males.
boys, respectively, the scale scores were as follows: BDI
The prevalence of mental disorders among controls was 18%.
score 7.0 (SD 5.7) vs. 3.0 (SD 2.6); YSR total score 47.7
Kim et al recently reported one in five (19%) community-
(SD 4.9) vs. 37.3 (SD 26.0); YSR internalizing subscale
dwelling children and adolescents aged 6–17 years to
score 13.7 (SD 2.5) vs. 11.7 (SD 7.1); YSR externalizing sub-
have mental health problems [25]. The similarity of these
scale score 17.7 (SD 1.5) vs. 9.7 (SD 9.9); and AUDIT score
two figures supports our finding that mental disorders are
5.7 (SD 8.1) vs. 1.0 (SD 1.7). Among self-cutting adolescents
nearly four times more common in currently self-cutting
both the BDI scores (p ¼ .05) and the YSR internalizing
adolescents than among other adolescents.
subscale scores (p ¼ .04) were higher in girls than in boys.
In Finland, illicit drug use is still less common than in
Finally, several logistic regression models were con-
many other Western countries. Heavy and frequent alcohol
structed. Boys were excluded because of their small number
use among adolescents is, however, of great concern from
in the study sample. Age, AUDIT score, YSR internalizing
the public health perspective. In this study the AUDIT score,
and externalizing subscale scores, and presence of major
which is an indicator of alcohol use disorders, was indepen-
depressive disorder and anxiety disorders were included in
dently associated with the presence of self-cutting. There
the final model (Table 3). The AUDIT score (p ¼ .016),
may be two explanations for this. Alcohol is known to
YSR internalizing subscale score (p ¼ .001), and major
have a rapid anxiolytic effect but, when used frequently, it
depressive disorder (p ¼ .04 8) were the factors that were
may provoke anxiety and depression. Anxious and depres-
independently associated with the presence of self-cutting
sive youngsters who have the experience of self-cutting
in girls. The model correctly classified 92% of cases.
may use alcohol as a means of self-help, but in the long
term this may exacerbate their symptoms and potentially
Discussion
Table 3
In this study we investigated the frequency of mental Crude and adjusted odds ratios with 95% confidence intervals for self-cutting
disorders in a sample of self-cutting, community-dwelling in girls
adolescents. Factors associated with self-cutting were also
Characteristic Crude OR (95% CI) Adjusted OR (95% CI)
explored. According to the results DSM-IV mental disorders,
especially major depressive disorder, are extremely common Age 1.00 (.74–1.35) 1.06 (.63–1.78)
among those community-dwelling girls who report self- YSR internalizing subscale 1.26 (1.15–1.39) 1.28 (1.11–1.47)
YSR externalizing subscale 1.18 (1.08–1.27) .95 (.83 –1.09)
cutting. Moreover, internalizing behavior and signs of alco- AUDIT score 1.25 (1.10–1.43) 1.31 (1.05–1.63)
hol misuse predict self-cutting independently of the presence Major depressive disorder 38.80 (7.13–160.31) 9.85 (1.02–94.72)
of mental disorders. Any anxiety disorder 4.15 (1.34–12.87) .49 (.05–5.08)
Major depressive disorder, anxiety disorders, and eating AUDIT ¼ Alcohol Use Disorders Identification Test; CI ¼ confidence
disorders were more common among adolescents who interval; OR ¼ odds ratio; YSR ¼ Youth Self-Report for adolescents aged
engaged in self-cutting than among controls. All of these 11–18 years.
 J. Hintikka et al. / Journal of Adolescent Health 44 (2009) 464–467 467

increase the need to repeat self-cutting for symptom allevia- [4] Haw C, Hawton K, Houston K, Townsend E. Psychiatric and personality
tion. Second, alcohol increases the risk of impulsive behav- disorders in deliberate self-harm patients. Br J Psychiatry 2001;178:48–54.
[5] Rodham K, Hawton K, Evans E. Deliberate self-harm in adolescents:
ior, which may lead to impulsive acts of self-cutting.
The importance of gender. Psychiatr Times 2005;22:36–40.
Internalizing behavior was independently associated with [6] TuiskuV Pelkonen M, Karlsson L, et al. Suicidal ideation, deliberate
self-cutting. Girls are more prone than boys to internalizing self-harm behaviour and suicide attempts among adolescent outpatients
[14,15], and girls in this sample also had higher subscale and depressive mood disorders and comorbid axis I disorders. Eur
scores for internalizing. Moreover, girls report intrapsychic Child Adolesc Psychiatry 2006;15:199–206.
[7] Zlotnick C, Mattia JI, Zimmerman M. Clinical correlates of self-muti-
reasons for self-harm more often than boys [26]. This finding
lation in a sample of general psychiatric patients. J Nerv Ment Dis 1999;
suggests that girls who direct their feelings inward have a risk 187:296–301.
of seeking alleviation for their symptoms of depression and [8] Ross S, Heath N. A study of the frequency of self-mutilation in a com-
anxiety by self-cutting. munity sample of adolescents. J Youth Adolesc 2002;31:61–77.
The small sample size is a major limitation of this study. [9] Martin G, Rozanes P, Pearce C, Allison S. Adolescent suicide, depres-
sion and family dysfunction. Acta Psychiatr Scand 1995;92:336–44.
Moreover, adolescents less than 15 years of age needed
[10] Muehlenkamp JJ, Gutierrez PM. An investigation of differences be-
written consent from parents before participating, which tween self-injurious behavior and suicide attempts in a sample of ado-
reduced the participation rate. It is possible that parents lescents. Suicide Life Threat Behav 2004;34:12–23.
who have fewer worries about their children more readily [11] Brunner R, Parzer P, Haffner J, et al. Prevalence and psychological cor-
give consent in this kind of study. This may have led to relates of occasional and repetitive deliberate self-harm in adolescents.
Arch Pediatr Adolesc Med 2007;161:641–9.
underestimation of the frequency of both current self-cutting
[12] Haavisto A, Sourander A, Multimaki P, et al. Factors associated with
at baseline and of mental disorders in this study. However, ideation and acts of deliberate self-harm among 18-year-old boys. A
the homogenous sample of community-dwelling adolescents prospective 10-year follow-up study. Soc Psychiatry Psychiatr Epide-
is a strength, as is the use of structured clinical interviews. miol 2005;40:912–21.
The diagnostic interviews were not conducted with inter- [13] Portzky G, van Heeringen K. Deliberate self-harm in adolescents. Curr
Opin Psychiatry 2007;20:337–42.
viewers blinded to cutting status, which might have biased
[14] Crick NR, Zahn-Waxler C. The development of psychopathology in fe-
the diagnoses towards greater severity in the cutting group. males and males: Current progress and future challenges. Dev Psycho-
Nevertheless, this is not likely to be a major source of error be- pathol 2003;15:719–42.
cause the preponderance of the self-reports also suggests that [15] Leadbeater BJ, Blatt SJ, Quinlan DM. Gender-linked vulnerabilities to
the cutters were more disturbed. Moreover, we do not have in- depressive symptoms, stress, and problem behaviors in adolescents.
J Res Adolesc 1995;5:1–29.
terrater reliability data for the diagnostic interviews; however
[16] Evans E, Hawton K, Rodman K, Deeks J. The prevalence of suicidal
round-table discussions to reach a consensus on diagnoses phenomena in adolescents: A systematic review of population-based
partially compensated for this shortcoming. Finally, we may studies. Suicide Life Threat Behav 2005;35:239–50.
have missed other factors that could be associated with self- [17] Favazza A. Why patients mutilate themselves. Hosp Comm Psychiatry
cutting, such as impulsivity and other aspects of personality. 1989;40:137–44.
[18] Beck AT, Ward CH, Mendelson M, Mock J, Erbaugh J. An inventory
Bipolar disorder was not found in this study among self-
for measuring depression. Arch Gen Psychiatry 1961;4:561–71.
cutting adolescents. However, the childhood and adolescent [19] Beck AT, Steer RA, Garbin MG. Psychometric properties of the Beck
onset of bipolar disorder may be as high as 61% among adult Depression Inventory: Twenty-five years of evaluation. Clin Psychol
bipolar outpatients [27]. It is therefore possible that some of Rev 1988;8:77–100.
the adolescents in this sample who received a diagnosis of [20] Saunders JB, Aasland OG, Babor TF, et al. Development of early de-
tection of persons with harmful alcohol consumption—II. Addiction
major depressive disorder will later switch to bipolar disor-
1993;88:791–804.
der. However, it is impossible to predict the number of cases [21] Knight JR, Sherritt L, Harris SK, et al. Validity of brief alcohol screen-
that may switch. Axis II borderline personality disorder ing tests among adolescents: A comparison of the AUDIT, POSIT,
might be another possible future diagnosis among self- CAGE, and CRAFFT. Alcohol Clin Exp Res 2003;27:67–73.
cutting adolescents. [22] Achenbach TN, Rescorla LA. Manual for the ASEBA School-Age
Forms & Profiles. Burlington, VT: University of Vermont, 2001.
In conclusion, current self-cutting is a strong indicator of
Research Center for Children, Youth & Families.
current mental disorders, especially major depressive disor- [23] First MB, Spitzer RL, Gibbon M, Williams JBW. Structured Clinical
der. Whenever self-cutting is identified, it should be followed Interview for DSM-IV-TR Axis I Disorders, Research Version, Patient
by psychiatric assessment with precise diagnostics. Edition. (SCID-I/P) New York: Biometrics Research, New York State
Psychiatric Institute, 2002.
[24] Kumar G, Pepe D, Steer RA. Adolescent psychiatric inpatients’ self-
References reported reasons for cutting themselves. J Nerv Ment Dis 2004;
192:830–6.
[1] Klonsky ED, Oltmanns TF, Turkheimer E. Deliberate self-harm in non- [25] Kim HK, Viner-Brown SI, Garcia J. Children’s mental health and fam-
clinical population: Prevalence and psychological correlates. Am J Psy- ily functioning in Rhode Island. Pediatrics 2007;119:S22–8.
chiatry 2003;160:1501–8. [26] Laye-Gindhu A, Schonert-Reichl KA. Nonsuicidal self-harm among
[2] Favazza AR. Bodies Under Siege. Self-mutilation and Body modifica- community adolescents: Understanding the ‘‘whats’’ and ‘‘whys’’ of
tions in Culture and Psychiatry. 2nd edition. Baltimore and London: The self-harm. J Youth Adolescence 2005;34:447–57.
John Hopkins University Press, 1996. [27] Post RM, Luckenbaugh DA, Leverich GS, et al. Incidence of child-
[3] Ghaziuddin M, Tsai L, Naylor M, Ghaziuddin N. Mood disorders in hood-onset bipolar illness in the USA and Europe. Br J Psychiatry
a group of self-cutting adolescents. Acta Paedopsychiatr 1992;55:103–5. 2008;192:150–1.