|
Completed Suicide Among Adolescents With No Diagnosable Psychiatric Disorder by Mauri J. Marttunen
Completed suicide for all ages is strongly associated with mental disorders (Black & Winokur, 1990). In general, the more thorough the data collection, the higher the reported proportion of suicide victims with mental disorders. In studies of adolescent suicides based on report data (e.g., coroners' records), the frequencies of depressive symptomatology, substance abuse, or general psychiatric symptomatology have ranged from approximately 50% to 90% of the cases (Cosand, Bourque, & Kraus, 1982; Graham & Burvill, 1992; Hoberman & Garfinkel, 1988; Poteet, 1987; Thompson, 1987).
The highest rates of mental disorders in representative samples of adolescent suicides have been reported in studies applying the psychological autopsy method (including informant interviews). Even in these studies, however, no evidence of any psychiatric disorder was found in 5% to 10% of the adolescents (Brent et al., 1988a; Marttunen et al., 1991; Shaffer et al., 1996; Shafii et al., 1988).
The influence of psychosocial stressors on suicidal behavior among adolescents has been emphasized (Brent et al., 1993b; Garrison et al., 1991; Shafii et al., 1988). The most commonly reported precipitating events in adolescent suicide have been interpersonal conflict, separation, rejection, and problems with discipline or the law (Brent et al., 1988a; Marttunen et al., 1993; Rich et al., 1990). Humiliating or shameful events may be particularly important as suicide precipitants (Apter et al., 1993; Shaffer, 1974).
Brent and associates (1993c) reported that adolescent suicide victims (under age 20) with no apparent psychopathology less often had histories of psychiatric treatment, suicide attempts, and other family members with affective illnesses, and had fewer interpersonal conflicts and total life stressors during their last 12 months, as compared with adolescent suicide victims with a psychiatric disorder.
However, in comparison with a community control group with no mental disorders, suicide victims with no apparent psychopathology showed some evidence of psychiatric risk factors.
Marttunen and associates (1994a) reported that 13- to 19-year-old adolescent suicide victims with adjustment disorders (including three without any disorders) seldom utilized psychiatric services but often had talked to somebody about their suicidal thoughts prior to committing suicide.
The present study analyzed data on all adolescent suicide victims in Finland during a 12-month period (see Marttunen et al., 1995). It sought to determine whether adolescents with no diagnosable psychiatric disorder differed from those with a psychiatric disorder (DSM-III-R axis I or II) in terms of family background, clinical characteristics, and suicide precipitants.
--------------------------------------------------------------------------------
METHOD
In the research phase of the National Suicide Prevention Project, all suicides (N = 1,397) committed in Finland between April 1, 1987, and March 31, 1988, were comprehensively analyzed using the psychological autopsy method (Clark & Horton-Deutsch, 1992; Litman et al., 1963). The determination of suicide was based on the results of the medicolegal examination required by Finnish law for all unnatural and unexpected deaths. During the research period, there were 116 suicide victims aged 13 to 22 years; 19 were females and 97 were males (Marttunen et al., 1995). For 5 males, the data were insufficient to make a psychiatric diagnosis, and these cases were excluded from the present analyses. This study examined male adolescent suicides: 8 with no diagnosable psychiatric disorder and 84 with a psychiatric disorder.
--------------------------------------------------------------------------------
Data Collection
A structured, face-to-face interview was conducted with at least one family member in 78 (93%) of the cases with a diagnosable psychiatric disorder and 7 (88%) of the cases without. The relatives were contacted by telephone or letter about one month after the suicide. The interview process always began by fully explaining the nature and procedures of the project, and informed consent was always requested and documented (Isometsa et al., 1996). Ethical aspects of the project were supervised by both the National Board of Health and the Ethics Committee of the National Public Health Institute.
The interview included 234 items concerning the victims' everyday life, family factors, use of alcohol and drugs, previous suicidality, help-seeking behavior, and life events. Interviews were usually conducted in the families' homes. The mean interval between the suicide and the interview was 137 days (SD = 92 days) for the psychiatric disorder group and 136 days (SD = 67 days) for the nondisorder group. The mean duration of each interview was 171 minutes (SD = 68 minutes) for the disorder group and 172 minutes (SD = 38 minutes) for the nondisorder group.
Health care professionals who had treated the suicide victims during the preceding year were interviewed face-to-face using a structured questionnaire containing 113 items dealing with health, treatment received, psychosocial symptoms, stressors, and level of functioning. Interviews were conducted for 17 (20%) of the disorder cases; none from the nondisorder group had had such contacts during the previous year. A semistructured 8-item interview was conducted either face-to-face or by telephone with the health or social care professional the suicide victim had last consulted. This interview was done for 2 (25%) of the nondisorder cases and 45 (54%) of those with a psychiatric disorder; the remainder usually had had no contact with a health care or social agency in the year preceding the suicide.
Supplementary unstructured interviews were conducted when needed, most often by telephone; this included 5 (63%) of the nondisorder cases and 48 (57%) of the disorder cases. At least one informant was interviewed (either a structured or supplementary interview) for all 8 of the nondisorder group and for 82 (98%) of those with a diagnosable psychiatric disorder.
Psychiatric and medical records, and the records of social service agencies and schools, as well as other reports, were available in 5 (63%) of the nondisorder cases and 67 (80%) of the disorder cases. Data from police and forensic reports were both available for the entire sample. Toxicologic analyses were available for all nondisorder cases and 78 (93%) of the disorder cases. Suicide notes were gathered for 1 (13%) nondisorder case and 31 (37%) of the disorder cases. At the conclusion of data collection, a multidisciplinary team discussed the cases, and comprehensive reports were written on the basis of all available information.
--------------------------------------------------------------------------------
DSM-III-R Diagnoses
Psychiatric diagnoses using DSM-III-R (American Psychiatric Association, 1987) were made in two phases. First, two psychiatrists independently made provisional best-estimate diagnoses. The act of suicide was not used as a diagnostic criterion. The interrater reliability of the provisional diagnoses was moderate to good (kappa = .50-.91) for the major diagnostic categories (Marttunen et al., 1995). The kappa coefficient for the "no diagnosable psychiatric disorder" category was .68 (95% cl., .41-.95). All cases with any diagnostic disagreement were reanalyzed with a third psychiatrist to achieve consensus for the final diagnoses. Multiple diagnoses on Axes I-III were allowed (Henriksson et al., 1993; Marttunen et al., 1991).
The suicide victims' psychological, social, and occupational functioning was assessed using the DSM-III-R Global Assessment of Functioning Scale (GAF). The assessment was based on the highest level of functioning for at least a few months during the past year. The one-week period prior to the suicide reflected most recent level of functioning. These assessments were based on consensus between two researchers.
--------------------------------------------------------------------------------
Suicide Precipitants
Psychosocial stressors were assessed using a life-event questionnaire administered to relatives (Heikkinen et al., 1992) and by a thorough review of all other data. When it was evident that a stressor occurring one month prior to the suicide had been a direct contributor, it was classified as a precipitant (Marttunen et al., 1993). The assessment of stressors and precipitants was based on the consensus of two of the investigators. Interpersonal precipitants included separation from, death of, or conflict with a parent, girlfriend or boyfriend, or other significant persons.
--------------------------------------------------------------------------------
Statistical Methods
The 8 male adolescents without a diagnosable psychiatric disorder were compared with the 84 males with a disorder in terms of suicide methods, family background, psychosocial stressors, previous suicidal behavior, and other appropriate clinical variables. Fisher's exact test (two tailed) and the Mann-Whitney U test were used (a .05 significance level was applied).
--------------------------------------------------------------------------------
RESULTS
Demography and Family Background
There were no statistically significant differences between suicide victims with and without a diagnosable psychiatric disorder by age (mean = 19.4 years, SD = 2.0, for the disorder group; mean = 18.4 years, SD = 2.8, for the nondisorder group) or parental socioeconomic status (for 75% of both groups, parents were manual or lower-level workers). None of the nondisorder group was found to have had a parent with alcohol problems or a psychiatric disorder, or who displayed violent or suicidal behavior. Compared with the disorder group, they tended to have experienced inadequate support from parents less often (see Table 1).
--------------------------------------------------------------------------------
Clinical Characteristics
None of the nondisorder group had made previous suicide attempts. Expressing suicidal thoughts was common in both groups, but those in the nondisorder group more often had communicated suicidal intent for the first time the day before the actual suicide (see Table 1).
--------------------------------------------------------------------------------
Axis III diagnoses were almost as common in the nondisorder group as they were in the disorder group. As expected, the level of psychosocial functioning during the previous year was higher for the nondisorder group than for the disorder group (mean GAF = 86.9, SD = 3.7, vs. mean GAF = 67.7, SD = 14.4, respectively; Mann-Whitney U, p [less than] .001). This also held true for the week preceding the suicide (mean GAF = 84.4, SD = 6.8, vs. mean GAF = 52.0, SD = 13.7, respectively; Mann-Whitney U, p [less than] .001). None of the nondisorder group had utilized psychiatric services, and contacts with other helping agencies also were more common in the disorder group.
Table 1. Background of 8 adolescent male suicides without and 84 adolescent male suicides with a current psychiatric disorder
* To see Table 1, go to the link below.
a) Parental alcohol abuse, severe parental somatic or psychiatric illness, parental attempted or completed suicide, or parental violence during past year
b) Recurrent truancy, recurrent stealing, recurrent running away from home, use of illicit drugs, sexual promiscuity, violence against person, arrests and convictions
1) Fisher exact test p = .07; 2) Fisher exact test p [less than or equal to] .05; 3) Fisher exact test p = .002
--------------------------------------------------------------------------------
Violent methods of committing suicide were common. The use of firearms (shooting) tended to be more prevalent in the nondisorder group (see Table 2). The disorder group more often committed suicide while under the influence of alcohol than did the nondisorder group. Of those who shot themselves, 2 out of 6 (33%) in the nondisorder group and 19 of 34 (56%) in the disorder group did so while under the influence of alcohol.
Suicide Precipitants
The mean number of psychosocial stressors during the month prior to the suicide was lower in the nondisorder group than in the disorder group (mean = 1.0, SD = 0.5, vs. mean = 3.1, SD = 1.5, respectively; Mann-Whitney U, p [less than] .001). The nondisorder group more often had difficulties with discipline during the final 24 hours than did those in the disorder group (see Table 3).
Table 2. Suicide methods and influence of alcohol in 8 adolescent male suicides without and 84 adolescent male suicides with a current psychiatric disorder
[TABULAR DATA FOR TABLE 3 OMITTED]
None of the nondisorder group had been unemployed or had problems at school or work during the final year, as compared with 32 (38%; Fisher's exact test, p [less than] .05) and 35 (42%; Fisher's exact test, p [less than] .03), respectively, of the disorder group.
--------------------------------------------------------------------------------
DISCUSSION
Consistent with previous findings (Brent et al., 1993c; Shaffer et al., 1996), 8% of the male adolescent suicide victims in Finland during a 12-month period had no diagnosable psychiatric disorder according to retrospective best-estimate assessments. The one exception to this rather uniform picture is the study by Apter et al. (1993) on Israeli male conscripts, which reported that 19% of the suicide victims had shown no evidence of Axis I psychopathology. However, preinduction screening was involved; thus subjects with severe psychopathology were excluded from that study's sample.
Family Environment
Family adversity is reportedly associated with youths' suicidal behavior (Pfeffer, 1989; Andrews & Lewinsohn, 1992). In the present study, severe parental psychopathology was reported in only the psychiatric disorder group. This finding is in agreement with Brent and associates (1993c), who found that suicide victims with no apparent psychopathology had lower rates of psychiatric illness and substance abuse in first-degree relatives as compared with suicide victims with a psychiatric disorder.
In studies such as the present one, where parents were interviewed after the suicide of their offspring, no systematic data on the quality of the interaction between the victims' family members can be obtained, and conclusions should be drawn with caution. Nevertheless, it seems that suicide victims with no diagnosable psychiatric disorder come from less disturbed family backgrounds than do those with a disorder.
Short Suicidal Process/Precipitating Disciplinary Problems
In agreement with previous research (Brent et al., 1993c), about a third of those with a psychiatric disorder had made previous suicide attempts, as compared with none of those in the nondisorder group. The prevalence of suicide talk in the nondisorder group (50%) was comparable to the prevalence of suicidal ideation (past week) among suicide victims with no apparent psychopathology (43%) in the study by Brent et al. (1993c).
They found suicidal ideation to be more common in those with a psychiatric disorder, whereas in the present study there was no marked difference in talking about suicide between the two groups. There was, however, a difference in the timing of suicide talk; the nondisorder group often communicated their thoughts just before the suicide, suggesting a shorter suicidal process. The unexpectedly high level of psychosocial functioning of the nondisorder group, even during the final week, may be another indication of a short suicidal process. Runeson (1992) also found that adolescents and young adults without psychiatric contacts had committed suicide after a short suicidal process.
Suicidal ideation is relatively common in adolescents (Fergusson & Lynskey, 1995), and thus a rather unspecific sign of suicide risk. Nevertheless, the finding that half of the nondisorder group did communicate their intentions suggests that overt suicidal communication, whether or not concurrent with evident psychopathology, should be viewed as a clinical warning sign. In one study (Brent et al., 1993a), suicidal ideation with a plan was as strongly associated with completed suicide as was a past suicide attempt.
The high frequency of firearms as the method of suicide among the nondisorder group, although not statistically different from that of the disorder group, supports the conclusion by Brent et al. (1993c) that restricting adolescents' access to lethal weapons may help to prevent suicides among adolescents with no psychiatric disorder.
Antisocial traits and difficulties with discipline or the law have been reported to correlate with adolescent suicidal ideation, suicide attempts, and completed suicide (Kashani et al., 1989; Apter et al., 1988; Shaffer, 1974; Marttunen et al., 1994b). In the present study, none of the nondisorder group had demonstrated antisocial behavior, but suicide precipitants did include problems with discipline. It seems that some adolescent males, facing disciplinary problems that cause acute feelings of shame or humiliation, may act rashly.
Methodology
The psychological autopsy is often the only way to study suicide victims' characteristics, and methodological studies have supported its use (Beskow & Runeson, 1991; Brent et al., 1988b). Nevertheless, the limitations of postmortem psychological studies of suicide involve the reliance on retrospective data provided by interviews of informants close to the victim. Such data are subject to recall error and also may be incomplete. Use of standardized assessments, several informants, and supplementary data from records helps to minimize these methodological problems (Clark & Horton-Deutsch, 1992).
Some differences between the compared groups may have gone undetected due to the relatively small number of suicide victims with no diagnosable psychiatric disorder. The possibility that some in the nondisorder group had hidden their psychiatric symptoms, or that the informants had not recognized them, cannot be ruled out. Further, because those in the nondisorder group had no psychiatric contacts, data from interviews of psychiatric professionals or from psychiatric records were not available, which may have had an impact on diagnostic classification. However, there was no statistically significant difference between the compared groups in either the number of interviewed informants or data from various records.
The strengths of the present study include the nationwide sample of adolescent suicides and the comprehensive data collection. The rate of family participation was high and, in most cases, several informants were interviewed and data from different records were accessed. Diagnostic classification included explicit criteria, and the final psychiatric diagnoses, as well as the assessment of psychosocial impairment and precipitants, were based on consensus.
--------------------------------------------------------------------------------
CONCLUSION
Suicide by adolescent males with no diagnosable psychiatric disorder seems to be a relatively impulsive act, occurring shortly after a precipitating event - such as the experience of loss or humiliation - and undertaken with an effective method. Due to the short suicidal process, relatively good psychosocial functioning, lack of diagnosable psychopathology, and infrequent contact with helping professionals, these suicides may be unexpected and there seems to be limited opportunities for preventive interventions. Communication of suicidal intent and precipitating problems with discipline appear to be among the few warning signs with these adolescents.
--------------------------------------------------------------------------------
COPYRIGHT 1998 Libra Publishers, Inc. in association with The Gale Group and LookSmart.
COPYRIGHT 2001 Gale Group
http://www.findarticles.com 
Back To The TopSMHAI Home |
About Suicide |
About Mental Health |
Suicide Prevention |
Suicide Survivors
Suicide Attempters |
Self-Injury - Cutters |
Crisis |
Donate |
SMHAI Library |
Online Support & Resources
Speakers & Presentations |
Memorials, Remebrances & Celebrations Of Life |
Healing Music
Suggested Reading - Survivors |
Suggested Reading - Attempters & Self-Injurers |
Mental Health Pros.
Upcoming Events |
Dr. Roerich's Welcome |
Ann Gay's Welcome |
Legal & About SMHAI
Privacy Policy |
Copyright Notice |
Awards Honoring SMHAI |
SMHAI Awards Program |
Contact
© SMHAI 2004 - 2006 All Rights Reserved.
No copying or redistribution without expressed written permission of SMHAI.
Logo Design by Allen R. Jacobson.
Site launched July 01, 2004.
|
