Relationship of intentional self-harm using sharp objects with depressive and dissociative tendencies in pre-adolescence–adolescence
Aims: The objectives of the present study were to (i) evaluate the prevalence of children and adolescents who have engaged in intentional self-harm using a sharp object; and (ii) investigate the relationship between self-harm with sharp objects and depressive tendencies or dissociative tendencies.
Methods: A total of 1938 students in grades 5–12 in Yokohama, Japan, were enrolled, and they completed anonymous self-report questionnaires including a question about intentional self-harm with a sharp object, the Depression Self-Rating Scale for Children (DSRSC) and the Adolescent Dissociative Experiences Scale (A-DES).
Results: The prevalence of self-harm using sharp object was 5.4% among male 5th–6th graders, 4.0% among female 5th–6th graders, 5.3% among male 7th–9th graders, 15.1% among female 7th–9th graders, 6.6% among male 10th–12th graders, and 9.6% among female 10th–12th graders. Categorical regression analysis showed that a small amount of variance in self-harm by sharp object was explained by DSRSC and A-DES scores.
Conclusions: Self-harm with a sharp object was prevalent among pre-adolescents and adolescents and was associated with depressive and dissociative tendencies.
INTENTIONAL SELF-HARM HAS been closely studied since the 1960s.1,2 In 1983 Pattison and Kahan reviewed 33 articles on deliberate self-harm on the body of low lethality and proposed a concept called deliberate self-harm syndrome.3 Hawton et al. reported that deliberate self-harm correlated with future suicidal behaviors.4
Briere and Gil conducted an epidemiological study on self-mutilation and reported that the prevalence of self-mutilation among the general population (18–90 years of age) was 4%, and the mean age at the first act of self-mutilation was approximately age 7.5 Evans et al. undertook a systemic review of 128 studies on suicidal phenomena in adolescents and concluded that the prevalence of deliberate self-harm during adolescence (12–20 years of age) was 13.2%.6
Matsumoto et al. studied 34 habitual self-mutilators, and indicated that self-mutilation closely correlated with depressive tendencies and dissociative tendencies.7 Sho et al. retrospectively studied 30 adolescent wrist-cutters and reported that 20% and 13% of adolescent wrist-cutters displayed mood and dissociative disorders, respectively.8
The previous observational studies suggest that the first act of self-harm typically occurs in pre-adolescence or adolescence,5,9,10 and depression and dissociation are related to self-harm in both adults and adolescents. Investigations of self-harm in pre-adolescents and adolescents that identify associated factors would provide important information for the early recognition of this behavior, which may lead to prevention of suicidal behaviors that can develop in the future. Few studies, however, have examined self-harm in pre-adolescence–adolescence, particularly in pre-adolescence. The objectives of the present study were to (i) evaluate the prevalence of children and adolescents who have engaged in intentional self-harm using a sharp object; and (ii) investigate the relationships between self-harm with a sharp object and depressive tendencies or dissociative tendencies.
Self-harm has been described in various terms, such as deliberate self-harm, self-mutilation, and self-injury. In the present study we defined self-harm as behaviors of an individual that intentionally harm their own body. The present study focused on self-harm using a sharp object, such as self-effected cutting, scratching and shaving using a sharp instrument, as defined in the International Statistical Classification of Diseases and Related Health Problems (10th revision; ICD-10), X78 (Intentional self-harm by sharp object).11
Subjects and procedure
Subjects consisted of 5th and 6th graders at five public elementary schools, 7th–9th graders at two junior high schools, and 10th–12th graders at one private high school in Yokohama. This survey was conducted from October 2007 to January 2008.
We selected 16 schools in Yokohama including 10 elementary schools, four junior high schools and two high schools with which we had been in contact through a workshop on children's mental health for teachers. We explained the purpose and methods to the principals of the schools, and requested them to participate in this survey. Consent was given by the principals of eight schools, consisting of five public elementary schools, two public junior high schools and one private high school. With regard to the private high school, the socioeconomic class of most of the families of the students was considered to be middle class. Many students enter this private high school after they graduate from public junior high school in Yokohama. Their academic levels are widely distributed, middle to high. Because we presumed that there was a little difference in the socioeconomic status and academic level between public elementary/junior high schools and a private high school, we compiled the data from these eight schools.
We delivered the survey sheets, return envelopes, and three different documents explaining the survey to the students, their parents and their teachers, respectively. The teachers distributed the document for the parents to the students and told them to give the documents to their parents. The document for the parents clearly stated that they could tell the teacher if they did not want their child to participate in this survey. Several days later, teachers distributed the document for the students explaining the survey to the students, along with the survey sheet and return envelope. The document for the students stated that they should answer the questions anonymously and submit the survey sheet in a sealed envelope to their teacher, if they were willing to consent to participate in this survey. The document also stated that they could return the survey sheet in a sealed envelope without answering any questions, if they did not wish to participate in the survey. Therefore, we regarded the collected survey sheets with answers as obtaining the consent of the students and their parents. Consent by signature was not obtained. The documents also stated the following: (i) their answers would not be disclosed to their teachers; and (ii) they could contact one of the authors if they had mental health difficulties. Students could fill out the survey sheets either at home or at school. This study protocol was approved by the ethics review board of Kanagawa Children's Medical Center.
Of the 3014 students (362 male, 342 female 5th and 6th graders; 586 male, 487 female 7th–9th graders; 605 male, 632 female 10th–12th graders) who received the survey sheets, 2498 students (82.9%) completed the survey (632 5th and 6th graders, 783 7th–9th graders, 1083 10th–12th graders). Among them, valid responses for all three questionnaires, that is, self-harm with a sharp object, Depression Self-Rating Scale for Children (DSRSC) and Adolescent Dissociative Experiences Scale (A-DES), were obtained from 1938 students (64.3%; 238 male, 247 female 5th and 6th graders; 305 male, 319 female 7th–9th graders; 362 male, 467 female 10th–12th graders). The present study included 1938 students for the analysis.
If we found a serious message, such as ‘I want to die’, we contacted the school. No participants needed psychiatric consultation.
Original question related to self-harm with a sharp object
The reliability and validity of this question has not yet been established. To the question, ‘Have you intentionally harmed your own body using a sharp or pointed instrument?’, students were asked to choose the best answer from the following four choices: response A: no self-harm, no thoughts of self-harm; response B: no self-harm, but thoughts of self-harm; response C: self-harm only once; and response D: self-harm more than once.
Depression Self-Rating Scale for Children
The DSRSC consists of 18 items related to childhood depression.12 Subjects were asked to answer each question on a 3-point scale (0, 1 or 2 points) based on experiences over the past week. The maximum score was 36 points. In Japan, Murata et al. prepared the Japanese version of the scale.13 The reliability and validity of the Japanese version of the DSRSC have been confirmed. They reported the validity using a cut-off of 16 between depressed and non-depressed children in Japan. The DSRSC has been found to be useful in children with ages ranging from 7 to 13 years. Recent studies have found that the DSRSC can be used on adolescents.14
Adolescent Dissociative Experiences Scale
The A-DES15 was prepared by modifying the Dissociative Experiences Scale (DES)16 for use in adolescents. The reliability and validity of the A-DES have been confirmed.15,17 The A-DES consists of 30 questions related to dissociative experiences, and subjects answered each question on an 11-point scale (0–10 points). The mean A-DES score for dissociative disorder patients has been reported at 4.85 ± 1.14.15 In Japan, Matsumoto et al. translated the scale into Japanese and reported high internal consistency (Cronbach's α = 0.944) and cross-validity.7,18
Statistical analysis was performed using SPSS version 16.0J for Windows Base System and SPSS Categories Option (SPSS, Chicago, IL, USA). Significance level was set at P < 0.05 (two-tailed test).
Response to self-harm with a sharp object
Subjects were divided into six groups according to sex (male and female) and grade (elementary school, 5th–6th; junior high school, 7th–9th; and high school, 10th–12th). Table 1 lists the ratio of each response regarding self-harm with a sharp object for each group. Among the six groups, the prevalence of past self-harm with sharp object was highest at 15.1% for female 7th–9th graders. The prevalence of self-harm was higher in female than male students among 7th–12th graders, and slightly higher in male than female students among 5th–6th graders. The prevalence of self-harm for 5th–6th graders was 4.8%, approximately half of that for 7th–9th and 10th–12th graders: 10.3% and 8.3%, respectively.
Table 1. Responses to the question about self-harm with a sharp object (%)
|A. No self-harm, no thoughts of self-harm||87.0||82.6||84.7||85.9||67.4||76.4||80.7||77.7||79.0|
|B. No self-harm, but thoughts of self-harm||7.6||13.4||10.5||8.9||17.6||13.3||12.7||12.6||12.7|
|C. Self-harm only once||2.5||2.0||2.3||2.3||6.0||4.2||3.6||3.6||3.6|
|D. Self-harm more than once||2.9||2.0||2.5||3.0||9.1||6.1||3.0||6.0||4.7|
DSRSC and A-DES scores according to sex and grade
Table 2 lists the mean DSRSC and A-DES scores versus sex and grade. Two-way analysis of variance was conducted to evaluate differences in mean DSRSC scores by sex and grade. The results showed a significant main effect for both sex and grade with a significant interaction.
Table 2. Mean DSRSC and A-DES scores vs sex and grade
|Depression self-rating scale for children†||7.7 ± 5.0||9.4 ± 6.3||10.6 ± 5.7||12.4 ± 6.9||13.5 ± 6.0||13.1 ± 6.1|
|Adolescent dissociative experiences scale‡||1.3 ± 1.4||1.3 ± 1.5||1.5 ± 1.5||1.6 ± 1.7||1.8 ± 1.6||1.6 ± 1.5|
Two-way analysis of variance was conducted to evaluate differences in mean A-DES scores by sex and grade. The results showed a significant main effect for grade only; there was no significant interaction effect between sex and grade.
DSRSC and A-DES scores for responses to question about self-harm with sharp object
Subjects were divided into six groups according to sex and grade, and DSRSC and A-DES scores for each group were compared in relation to responses to the question about self-harm with a sharp object (Tables 3,4). In all groups the DSRSC and A-DES scores for response A were lower than those for responses (b–d). The DSRSC and A-DES scores for response B were comparable to those for response C.
Table 3. Mean DSRSC scores vs response to question about self-harm with sharp object
|A. No self-harm, no thoughts of self-harm||7.1 ± 4.7||8.0 ± 5.1||10.0 ± 5.4||9.9 ± 5.2||12.4 ± 5.7||11.9 ± 5.5|
|B. No self-harm, but thoughts of self-harm||10.9 ± 3.7||15.3 ± 7.1||15.4 ± 6.1||16.5 ± 6.7||18.1 ± 5.0||17.4 ± 5.9|
|C. Self-harm only once||10.3 ± 2.3||17.4 ± 2.6||13.0 ± 6.8||16.4 ± 6.3||18.2 ± 5.5||16.7 ± 5.8|
|D. Self-harm more than once||15.7 ± 7.6||21.2 ± 6.6||14.0 ± 5.0||20.0 ± 8.2||17.8 ± 7.7||19.1 ± 6.1|
Table 4. Mean A-DES scores vs response to question about self-harm with sharp object
|A. No self-harm, no thoughts of self-harm||1.1 ± 1.2||1.0 ± 1.2||1.3 ± 1.3||1.2 ± 1.2||1.5 ± 1.5||1.3 ± 1.2|
|B. No self-harm, but thoughts of self-harm||2.4 ± 2.1||2.5 ± 2.1||2.4 ± 1.6||2.1 ± 1.9||2.8 ± 1.7||2.6 ± 1.8|
|C. Self-harm only once||1.4 ± 1.2||2.7 ± 1.7||3.6 ± 2.2||2.4 ± 1.8||3.2 ± 2.1||2.7 ± 1.9|
|D. Self-harm more than once||2.6 ± 1.5||4.5 ± 2.3||3.3 ± 2.2||3.4 ± 2.3||3.5 ± 2.1||2.8 ± 2.0|
Categorical regression analysis
To investigate the relationship of sex, grade, depressive tendencies and dissociative tendencies to self-harm with a sharp object, categorical regression analysis was conducted using responses to the question about self-harm using sharp object as the dependent variables and sex, grade and DSRSC and A-DES scores as independent variables (Tables 5,6). Responses to the question about self-harm with a sharp object were analyzed on a 4-point spline ordinal scale. Sex was a nominal scale, and grade was a spline ordinal scale. DSRSC and A-DES scores were numerical scales, and analysis was performed by dividing the scores into seven groups in a normal distribution. Responses to the question about self-harm with a sharp object were quantified as shown in Table 5. Table 6 lists the results of categorical regression analysis. Self-harm with a sharp object exhibited significant correlations to all four factors: sex; grade; DSRSC; and A-DES. The explanation rate of the regression formula (adjusted R2) was 0.231. The standardized coefficient for DSRSC scores was 0.329, and DSRSC had the closest correlation. The standardized coefficient for A-DES scores was 0.226, the second highest among the four factors. These results of categorical regression analysis showed that a small amount of variance in self-harm with a sharp object was explained by DSRSC and A-DES scores.
Table 5. Responses to question about self-harm with sharp object
|A. No self-harm, no thoughts of self-harm||1543||−0.494|
|B. No self-harm, but thoughts of self-harm||239||1.668|
|C. Self-harm only once||67||1.668|
|D. Self-harm more than once||89||2.824|
Table 6. Categorical regression analysis
The present results showed that the prevalence of self-harm with a sharp object among 5th–6th graders was 4.8%, approximately half of that among 7th–12th graders. To the best of our knowledge, no previous studies have investigated the prevalence of self-harm among elementary school children. The present study showed that even elementary school children are at risk of self-harm.
Several studies have been conducted on junior high school and high school students using self-report questionnaires in Japan. The present results are consistent with those previous studies, which reported that the prevalence of self-injury in male and female junior high school and high school students (mean age, 14.7 ± 1.4 years) was 7.5% and 12.1%, and that the prevalence of self-cutting in male and female junior high school students was 8.0% and 9.3%, respectively.10,19 The present results also are consistent with the prevalence of self-harm in other countries: 8.5% of high school students in the UK (self-cutting) and 13.9% of Canadian adolescents. 9,20
According to Hawton et al., when conducting school-based surveys of this kind, an important subgroup of pupils will inevitably be excluded who might be more likely to have experiences of self-harm and severe psychopathology.21 There is a possibility that the present study is underestimating the prevalence of self-harm in the students who were excluded from this analysis.
In the present study the prevalence of self-harm with a sharp object was slightly higher in male 5th–6th graders than in female 5th–6th graders, and was higher in female 7th–9th graders than in male 7th–9th graders. The prevalence of self-harm was comparable between 5th–6th graders and 7th–9th graders among boys, whereas the prevalence of self-harm was significantly greater for 7th–9th graders than for 5th–6th graders among girls. The reason for this gender difference might be related to the differences in emotional development in early adolescence.
Many studies have identified a correlation between self-harm and a history of traumatic events, especially sexual and physical abuse.5,7 Studies have also shown that future risks of illicit drug use and suicidal behaviors are high for people with a history of self-harm.4,19 Therefore, the existence of self-harm among pre-adolescence and adolescence should be highlighted as a high-risk behavior. Educational and mental health professionals should observe these high-risk pre-adolescents and adolescents cautiously for self-harm.
In the present study, self-harm during pre-adolescence and adolescence was predicted by depressive and dissociative tendencies on categorical regression analysis. These findings matched the clinical observation of wrist-cutting during adolescence as reported by Sho et al.8
Many reports have discussed the correlation between self-harm and depressive mood.3,7,9,20 Nixon et al. reported that self-harm was used for affect regulation, and coping with feelings of depression was the most common answer for self-harm in hospitalized adolescents.22 In the present study DSRSC score for students with self-harm more than once was lowest for male 5th–6th graders at 15.7 ± 7.6 and highest for female 5th–6th graders at 21.2 ± 6.6. Because the cut-off for DSRSC scores in Japan is 16,13 the prevalence of pathological depressive tendencies was considered to be high among students who have performed self-harm more than once. Matsumoto et al. reported that the mean A-DES score of habitual self-mutilators was higher than that for controls, and considered habitual self-mutilation as possibly associated with a dissociative phenomenon.7 In the present study A-DES score for students with self-harm more than once was lowest for male 5th–6th graders at 2.6 ± 1.5 and highest for female 5th–6th graders at 4.5 ± 2.3. Based on A-DES score (4.85 ± 1.14) for dissociative disorder patients,15 the prevalence of pathological dissociative tendencies was considered to be high among students who have performed self-harm more than once. Self-harm, particularly repeated self-harm with a sharp object, seems to have a considerable indication of mental health problems. The present results suggest that we should examine pre-adolescents and adolescents with self-harm from the perspective of depression and dissociation.
In the present study, DSRSC and A-DES scores for students who had thought about self-harm but who had not engaged in this activity were higher than those for students who had not thought about self-harm, and were comparable to those for students who had performed self-harm only once. These findings suggest that the prevalence of mental health disorders is high in pre-adolescents and adolescents who think about self-harm, and preventative interventions are required for these pre-adolescents and adolescents in the field of mental health.
Various limitations were present in the current study. First, private high schools attract students from a wider geographic area, and a small bias might have existed in academic performance. In other words, high school students in the present study may not have fully represented the region of Yokohama. Second, elementary and junior high schools were not randomly selected, and regional representativeness was thus low. Third, the reliability and validity of the original question related to self-harm with a sharp object has not yet been established. Fourth, the present study used only self-report questionnaires without structured interviews, which was insufficient for objective assessment.
In the future, epidemiological study on a group with high regional representativeness is required. In addition, techniques such as structured interview are necessary to identify further background factors for self-harm.
We wish to express gratefulness to the students, their families and their teachers who consented to participate in the present study.