The
Relationship Between Adolescent Suicidal Behavior and Life Events in Childhood and Adolescence Erik J. de Wilde, M.Sc., Ineke C.W.M. Ren#{233} F.W. Diekstra, Ph.D., and Willem
Objective: Although the relationship between adolescent suicidal behavior has frequently been ies
oflife
events
have
been
initiated
that
and depressed adolescents. Therefore, with both depressed and nondepressed to life years)
events that happened and adolescence (age
Kienhorst, Ph.D., H.G. Wolters, Ph.D.
experience recognized
of problematic during the past
discriminated
between
adolescent
the authors adolescents
compared who never
adolescent attempted
in two periods: 1 2 and older).
childhood Method:
life events and decade, few studsuicide
attempters
suicide attempters suicide with respect
(defined as the period Using a semistructured
up to age 12 interview, the
authors gathered life event data about childhood and adolescence from three groups of adolescents: 48 suicide attempters, 66 depressed adolescents who had never made a suicide attempt, and 43 nondepressed adolescents who had never made a suicide attempt. Results: The group of adolescents who attempted suicide differed from both of the other groups in that they
had
during
experienced
last year to repeat
before
During the attempt,
a class,
embedded their
more
adolescence.
occurred.
turmoil
rooted
families,
they
further
instability,
social
Conclusions:
not just in the problems
families,
in their
adolescence,
in childhood
traumatic events during adolescence (Am J Psychiatry 1992; 149:45-51)
For
every and
and
social
Received Jan. 10, 1991; revision received June 5, 1991; accepted July 16, 1991. From the Department of Pediatrics, University of Utrecht, and the Department of Clinical, Personality and Health Psychology, University of Leiden, The Netherlands. Address reprint requests to Dr. de Wilde, Wilhelmina Children’s Hospital, Psychosocial Department, P.O. Box 18009, 3501 CA Utrecht, The Netherlands. Supported by the Praeventiefonds, the Programmeringscollege Onderzoek Maatschappelijke Dienstverlening (Ministry of Welfare, Health and Cultural Affairs), the K.F. Heijnfonds, the Stichting voor Vergeten Groepen, the Stichting Elise Mathilde Fonds, the Stichting “Dc Balije,” the Prof. Dr. Weijers Stichting, the Evert Zoudenbalchhuis, Crediet & Effectenbank, and the Stichting Lam#{233}ris Foundation. Copyright © 1992 American Psychiatric Association.
January
1992
suicidal
in childhood
often
such
sexually
as changes
adolescents, during
instability
and
not
abused.
the
suicide
stabilizing
During
in residence
has to deal with
not stabilizing
n their impressive review of attempted suicide in adolescence, Spinito et al. ( 1 ) stated, “Presence of stressful life events . . . seems to have important implications for understanding suicide attempts in children and adolescents.” Nevertheless, although the relationship between experience of problematic life events and adolescent suicidal behavior has frequently been recognized during the past decade (2, 3), few studies of life events have been initiated that discriminated between adolescent suicide attempters and depressed adolescents. This comparison is essential because many life
149:1,
more
adolescent
I
Am J Psychiatry
starting
were
and attempt
but in greater
adolescence, in the year
the
having seems
turmoil
in combination preceding
in with
the attempt.
events that take place in the lives of suicide attempters are also found in the lives of adolescents who report depressed mood (4), which implies that not taking depression into account is problematic when identifying life events specific for adolescent suicide attempters. Several studies compared suicidal adolescents with heterogeneous psychiatric groups of nonsuicidal adolescents (5-7) or with other clinical groups (8, 9). However, it is not clear from these studies whether the compared groups were depressed or hot.
Two
studies
(10,
1 1), however,
made
comparisons
using depressed groups. Friedman et al. (10) compared suicidal depressed adolescents with nonsuicidal depressed adolescents but restricted their research to family history of illness. Chronic psychiatric illness of a parent (especially depression) was more frequently found among suicide attempters. Cohen-Sandlen et al. (11) compared suicidal children and young adolescents (514 years old) with depressed and other psychiatric control groups regarding life events during different developmental periods. They found that more life events
were
related
to suicidal
behavior
If the results of Cohen-Sandler cable to adolescents, they would
with
increasing
age.
et al. were also appliagree with those of
45
ADOLESCENT
SUICIDAL
BEHAVIOR
Jacobs (12), who compared SO adolescent suicide attempters with matched control subjects from schools. He hypothesized that 1) the first of four phases leading to a suicide attempt consists of a longstanding history of problems from childhood to the onset of adolescence and 2) the second phase is a period of “escalation of problems” at the onset of adolescence in excess of those “normally” associated with adolescence (p. 64). In our opinion, Jacobs failed to find adequate support for his valuable hypotheses because his data analysis was rather obscure and because, by comparing suicidal adolescents with normal adolescents, he did not account for depression. A comparison of suicide attemptens with both depressed and “normal” adolescents (as a baseline for “normal” adolescent turmoil) may give more solid information about his hypotheses. In addition, analysis of Jacobs’ hypotheses requires the specification of the period (childhood or adolescence) during which life events have taken place. Therefore, the present study compared adolescent suicide attempters with both depressed and nondepressed adolescents who never attempted suicide with respect to life events that happened in two periods: childhood (defined as the period up to age 12) and adolescence (age 12 and older). Moreover, because of the growing evidence about the importance of recent life events negarding suicide attempts and psychiatric disorder in young people (1 1, 13, 14), the last year before the attempt (for the suicide attempters) on before the interview (for the other groups) was studied separately.
METhOD Subjects Three groups of subjects were asked to cooperate with this study. The first group comprised 41 girls and seven boys 14-21 years old (mean±SD age=17.S±1.S years) who attempted suicide 1-52 weeks before the measurement date. The second group consisted of 52 girls and 14 boys 14-21 years old (mean=17.9±1.6) who suffered from a depressed mood at the time of assessment and had never attempted suicide. (For reasons of readability, the adolescents who suffered from a depressed mood will be referred to as “depressed adolescents”
in the
remainder
of the
text.)
The
third
group
comprised 31 girls and 12 boys 14-21 years old (mean=17.S±1.3), randomly selected from a “normal” school population. The three groups were not matched, but there were no significant differences among them in relevant sociodemographic features such as age (F= 1.45, df=2, 154, p=O.24), sex (X2=2.43, df=2, p=O.3O), and current type of school (x2=7.22. df=6, p=O.l2). Two methods were used to recruit subjects for this study.
First,
institutions
for
mental
health
care
were
asked for referrals, and, second, 9,393 secondary school students were screened (iS). Approximately half of the suicide attempters and half of the depressed adolescents were recruited through the institutions of men-
46
tal
health
cane.
The
rest
of the
suicide
attempters
and
depressed adolescents as well as the entire group of normal adolescents were recruited from among the 9,393 secondary school students. The suicide attempters and the depressed adolescents did not differ significantly in this
recruitment
procedure
(2=3.1,
df=2,
p=O.O8).
The study was performed in compliance with the U.S. Human Subject Research Guidelines. Informed consent was obtained from all subjects and their parents. Almost two-thirds of the suicide attempters had taken an overdose of medicine on toxic substances, onequarter had injured themselves, and the rest had thrown themselves in front of a moving vehicle, tried to hang themselves, on jumped from a height. The mean±SD
scone
of the
suicide
attempters
on the
Suicide
Intent Scale (16), 6.8±3.4, was comparable to that of SO British adolescents who attempted suicide (2). Three suicide attempters had scores of 1 3, indicative of a strong suicide intent. The mean risk scone of all 48 suicide
attempters
8.1±2.1,
on
the
indicative
rescue
score
of
on this
Risk-Rescue
Rating (1 7) was risk, and their 1 1 .6±2.3, indicative of a
a moderately
scale
was
low
moderately high rescue level. The 66 depressed adolescents sion
at two
separate
were
times.
First,
tested
for depres-
at the
time
of the
screening procedure preceding the interview, they were selected randomly if they appeared to have a score of 46 on higher on the Zung Self-Rating Depression Scale ( 1 8) and had a score of at least 1 6 on the Depression Adjective Checklist (19). The depressed adolescents refenred from the institutions of mental health care met Research Diagnostic Criteria (20) for major depressive disorder. The second assessment took place at the time of the interview. Subjects who had a scone of 46 on higher on the Zung scale at that time were included in this study. Their mean scone on the Zung scale was 51.9±5.2, compared with 49.4±1 1 .7 for the suicide attempters and 33.5±5.6 for the normal subjects.
For the normal initially
comparison
selected
the secondary selection
school
were
these
made
a suicide
scone
higher
than
time
of the
interview,
and
43 adolescents Measures
and
Checklist.
were among
The criteria that
scale met
from
they on 16
had
did
not
on
the
at the
the criteria
the study,
comparison
for
they
Nevertheless,
students
in the normal Data
Zung
removed
from
reported
and
the
two
were
screened. students
on
45 students
interview
attempt
46
Adjective
depression
for
students
that
never
Depression
group,
at random
for
leaving
group.
Analysis
A semistructured interview including several self-neport questionnaires was conducted with each subject by a graduate psychology student or a psychologist who was especially trained for this purpose. Before the actual data collection, a pilot study was carried out with eight adolescents; the interview was modified on the basis of the
findings
of this
pilot
study.
The Life Event Time Schedule, a part of the Life History Interview (4), was developed for this study. The
Am J Psychiatry
149:1,
January
1992
DE WILDE,
TABLE 1. Number of Major Life Events Experienced Depressed Adolescents, and Normal Adolescents
During
Childhood
12
(up to
Suicide Attempters (N=48)
Years
of Age)
KIENHORST,
by Adolescents
Depressed Adolescents (N=66)
DIEKSTRA,
ET AL.
Who Attem pted Suicide,
Normal Adolescents (N=43)
Event
Mean
SD
Mean
SD
Mean
SD
KruskalWallis H
Serious illness Accident Change in living situation Change in caretaker Separation of parents Reunion of parents Change of residence Death of family member Physical health event of family member Mental health event of family member Unemployment of family member Imprisonment of family member Sibling leaving home Change of school or type of school Repeating a class Ending a relationship Physical abuse Sexual abuse Criminal events Totalnumberofevents
0.60 0.69 0.38 0.48 0.23 0.13 0.83 0.08 1.19 0.56 0.25 0.06 0.13 0.71 0.10 0.00 0.19 0.17 0.04 6.81
0.82 0.78 0.84 0.85 0.42 0.33 1.02 0.40 1.39 1.01 0.48 0.24 0.44 0.92 0.31
0.44 0.55 0.30 0.21 0.06 0.05 0.74 0.02 0.89 0.38 0.11 0.01 0.11 0.77 0.09
0.68 0.77 0.80 0.54 0.24 0.21 0.85 0.12 0.99 0.76 0.31 0.12 0.47 0.84 0.29
0.56 0.56 0.12 0.16 0.09 0.07 0.84 0.02 1.23 0.28 0.07 0.00 0.09 0.65 0.05
0.73 0.80 0.45 0.43 0.29 0.26 0.92 0.15 1.66 0.67 0.26
1.23 1.73 4.19 5.62 7.81a 2.51 0.23 0.84 0.46 2.61 5.83
-
4.03
0.37 0.72 0.21
0.21 0.63 1.08
-
0.00
-
0.00
0.14 0.05 0.12 5.03
0.43 0.21 0.41 3.68
0.00 0.05 0.19 5.02
-
0.00
0.49 0.48 0.29 3.91
significant significant significant
ape#{216}5;
bp