Nonsuicidal self-harm and suicide attempts in adolescents: differences in kind or in degree? Anita J. Tørmoen, Ingeborg Rossow, Bo Larsson & Lars Mehlum
Social Psychiatry and Psychiatric Epidemiology The International Journal for Research in Social and Genetic Epidemiology and Mental Health Services ISSN 0933-7954 Volume 48 Number 9 Soc Psychiatry Psychiatr Epidemiol (2013) 48:1447-1455 DOI 10.1007/s00127-012-0646-y
1 23
Your article is protected by copyright and all rights are held exclusively by SpringerVerlag Berlin Heidelberg. This e-offprint is for personal use only and shall not be selfarchived in electronic repositories. If you wish to self-archive your article, please use the accepted manuscript version for posting on your own website. You may further deposit the accepted manuscript version in any repository, provided it is only made publicly available 12 months after official publication or later and provided acknowledgement is given to the original source of publication and a link is inserted to the published article on Springer's website. The link must be accompanied by the following text: "The final publication is available at link.springer.com”.
1 23
Author's personal copy Soc Psychiatry Psychiatr Epidemiol (2013) 48:1447–1455 DOI 10.1007/s00127-012-0646-y
ORIGINAL PAPER
Nonsuicidal self-harm and suicide attempts in adolescents: differences in kind or in degree? Anita J. Tørmoen • Ingeborg Rossow Bo Larsson • Lars Mehlum
•
Received: 5 May 2011 / Accepted: 15 December 2012 / Published online: 27 December 2012 Ó Springer-Verlag Berlin Heidelberg 2012
Abstract Purpose The purpose of the present study was to measure the prevalence of self-harm (SH) behaviours and examine potential differences in characteristics among adolescents reporting on self-harm (SH), depending on whether they had attempted suicide (SA), performed nonsuicidal selfharm (NSSH), or both. Methods Cross-sectional survey of 11,440 adolescents aged 14–17 years in the city of Oslo, Norway. Responses regarding measures of lifetime SH and risk factors were collected. The response rate was 92.7 %. Data were analysed by segregating SH responses into the categories of NSSH, SA, and NSSH ? SA. Results Among all respondents, 4.3 % reported NSSH, 4.5 % reported SA, 5.0 % reported both NSSH and SA, and 86.2 % reported no SH. The group reporting to have engaged in both behaviours comprised more girls and reported more suicidal ideation, problematic lifestyles, poorer subjective health, and more psychological problems compared with the other groups. The four groups could be distinguished by one discriminant function that accounted for most of the explained variance.
A. J. Tørmoen (&) I. Rossow L. Mehlum National Centre for Suicide Research and Prevention, Institute of Clinical Medicine, University of Oslo, Sognsvannsveien 21, Building 12, 0372 Oslo, Norway e-mail:
[email protected] I. Rossow Norwegian Institute for Alcohol and Drug Research, POB 565 Sentrum, 0105 Oslo, Norway B. Larsson Department of Neuroscience, Regional Centre for Child and Adolescent Mental Health, NTNU, Trondheim, Norway
Conclusions Our findings suggest that NSSH and SA are parts of the same dimensional construct in which suicidal ideation carries much of the weight in adolescents from a school-based sample. They also indicate the group of adolescents who seems to alternate between NSSH and SA is more burdened with mental ill-health and behavioural problems compared with others. These adolescents should therefore be targeted by clinicians and school health personnel for identification and provision of adequate help and services. Keywords Self-harm Suicide attempts Adolescents General population
Introduction Suicide is a worldwide, major public-health issue. The most prominent predictor of suicide is previous engagement in self-harm (SH) [1–4], with up to 60 % of suicide cases having previously engaged in SH [5]. Adolescence is the peak period for SH [6, 7] and with a lifetime prevalence rate of SH *13 % [8]. SH in adolescents is more prevalent among girls and those with depressive symptoms, eating problems, substance abuse, antisocial behaviour, poor physical health, and low self-esteem [9–12]. These are indicators of serious distress which call for more research of self-harming behaviour in adolescents, not only as a search for precedents for suicide, but also in a public-health perspective. It is common to distinguish between SH with the intent to end one’s own life and self-harming behaviours with other intentions. The terminology in this area is, however, complex, and numerous terms have been applied [13]. In this study, we define self-harm (SH) as intentional selfpoisoning or self-injury, regardless of intention to die [14].
123
Author's personal copy 1448
This broad category is then further divided into ‘‘suicide attempt’’ (SA) referring to self-harm behaviour with an explicit or inferred intention to die [9] and ‘‘nonsuicidal self-harm’’ (NSSH) referring to self-harm without any suicide intent [13]. The definitions of SH were used in the Child and Adolescent Self-Harm in Europe (CASE) study, which in turn were based on definitions used in the large, multicenter WHO/EURO Study of parasuicide [15]. The European tradition of asking about intentional self-harm regardless of suicidal intent started two decades ago. With this study, we approach a tradition in USA, which distinguishes between self-harm episodes with and without suicidal intent. In USA, the former is referred to as suicide attempts and the latter as nonsuicidal self-injury. However, the term nonsuicidal self-injury excludes overdoses. Considering that overdose or at least minor overdose is a common method of self-harm [16], and not only seen in suicide attempts, the term nonsuicidal self-harm is applicable, as in published literature from Canada and Australia [17, 18].There is a dearth of studies on the similarities and differences between these various forms of self-harming behaviour. The present study is, to our knowledge, among the few large-scale community-sample studies to include questions about both these types of self-harming behaviour and to report in particular on the group who had engaged in both behaviours. The approaches to explain SA tend to extract from theories on completed suicide, whereas theories on emotional dysregulation are often applied to explain NSSH. To distinguish between SH behaviours with and without suicide intent is an important and complex task for clinicians. One such attempt to distinguish these behaviours seems to be based on the assumption that patients that engage in NSSH are more attention seeking and manipulative and less in genuine need of mental health care compared with patients that engage in SA. This dichotomization of the two behaviours could, however, lead mental health professionals to underestimate the severity and potential lethality of NSSH [19]. Instead, one could argue that the difference between the two types of self-harm is one of a degree rather than one of a kind. This is based on the knowledge that many of the same risk factors are shared and that, particularly among adolescents, suicide intent may fluctuate rapidly and the suicidal person may be highly ambivalent [6]. Alternatively one could view NSSH as a distinct clinical syndrome that should be diagnosed separately; such a diagnostic category has recently been proposed for the upcoming Diagnostic and statistical manual of mental disorders, fifth edition (DSM-V) [20]. Among the distinctive features of NSSH, it has been claimed that damage to the skin, which is likely to induce pain, is generally accompanied by a relief from negative feelings, or elicitation of pleasurable feelings or gratification [21].
123
Soc Psychiatry Psychiatr Epidemiol (2013) 48:1447–1455
However, according to the authors of the proposal for a new DSM-V diagnosis, there is a need for additional research addressing the similarities and differences among the various forms of self-harming behaviour among adolescents. Several studies have used data from clinical populations to show a significant overlap between SA and NSSH, i.e., a significant fraction of subjects who report NSSH will also report SA at some point in time [22, 23]. Conversely, a significant fraction of suicide attempters will also report NSSH [19, 24]. These studies suggest that suicide attempters and self-injuring adolescents do not constitute distinctive, but are rather partly overlapping, population groups. Furthermore, clinical studies indicate that adolescents who engage in both types of behaviour are more likely to report mental health problems, such as depression and anxiety, impulsiveness [19], and problem behaviour [24], compared with other SH patients. Thus, whether the various types of SH constitute distinctive behavioural categories or whether they should be regarded as parts of a behavioural spectrum and continuum of severity remains unclear. To date, nearly all of the knowledge available on the relationship between SA and NSSH in adolescents has been derived from clinical samples. Since, however, clinical studies may be strongly biased by referral and selection, studies based on community samples are needed to clarify whether knowledge derived from clinical studies is generalizable to the general population of adolescents. Four recent community studies on adolescents provided new data in this direction [22, 25–27]. However, these studies were either based on small samples or did not address the substantial overlap between the self-harming behaviours and the proportion of adolescents who had engaged in both NSSH and SA. Although NSSH and SA have been investigated in adolescents in previous research, few studies in community samples of adolescents reported specifically on the prevalence of adolescents having engaged in these behaviours separately or on the prevalence of having engaged in both. To the best of the authors’ knowledge, no large community-based sample studies have been published that directly address the relationship between NSSH, SA, and NSSH ? SA. The purpose of the present study was twofold: (1) to assess the prevalence of self-harming behaviours, in particular the combination of NSSH and SA behaviours among adolescents in the general population; and (2) to explore possible differences and similarities among these groups regarding demographic and psychosocial characteristics. Analyses compared those engaging in only one of the behaviours (SA or NSSH) with those engaging in both (NSSH ? SA) behaviours and contrasted them with adolescents with no SH. Based on previous studies performed in clinical settings, we hypothesized that, in a sample of
Author's personal copy Soc Psychiatry Psychiatr Epidemiol (2013) 48:1447–1455
adolescents from the general population, those exhibiting NSSH or SA would be fairly similar regarding psychosocial factors and that differences would primarily be found regarding suicidal ideation. Furthermore, we hypothesized that those engaging in both behaviours would constitute a subgroup with a particularly high risk of experiencing psychosocial problems compared with subjects who engaged in only one type of these behaviours.
Sample and methods Participants and procedures The present study was based on data from a large, comprehensive, and cross-sectional survey of adolescents in the city of Oslo, Norway. Data were collected in February and March 2006. All junior and senior high schools (n = 91) in the city were asked to participate in the study. Sixteen of the 91 schools declined to participate in the survey, for various reasons. The gross sample comprised all pupils in grades 9, 10, and 11 and the response rate in the participating 75 schools was 92.7 %. Geographically, the nonparticipating schools were distributed evenly in the city; thus, no socioeconomic sampling bias was expected. Study participation was based on informed passive parental consent. The net sample comprised 11,440 respondents with a mean age of 15.4 years (range 14–17 years) and 51.2 % of the sample were girls. The pupils completed a comprehensive questionnaire at school, during two school hours. Measures In the present study we have aimed to give the respondents a possibility to report SH separately on episodes of SA and NSSH. Self-harming behaviour and suicidal behaviour were assessed using two questions: (1) on self-harm (SH), ‘‘Have you ever taken an overdose of pills or otherwise tried to harm yourself on purpose?’’ (‘‘no’’, ‘‘yes, once’’, and ‘‘yes, more than once’’), which was a question derived from the CASE study and also used in several other studies [8, 25, 28, 29]; (2) on suicide attempt (SA), ‘‘Have you ever tried to kill yourself?’’ (‘‘no’’, ‘‘yes, once’’, and ‘‘yes, more than once’’), which is a question that was used in previous Norwegian studies of school adolescents [30]. On the basis of the responses to these two questions the pupils were subdivided further into one of four possible categories of SH: (1) no SH (a response of ‘‘no’’ to both questions); (2) NSSH (‘‘yes’’ to SH and ‘‘no’’ to SA); (3) SA (‘‘no’’ to SH and ‘‘yes’’ to SA); and NSSH ? SA (‘‘yes’’ response to both questions). Thus, we assumed that those who confirmed SH but not SA had no suicidal intent and therefore
1449
engaged in NSSH, and those who confirmed SA but not SH had engaged in suicidal attempts only. Previous studies have also categorized SH by intent, based on the same SH question that we used [17, 25]. Sociodemographic variables included information on gender, age, and living arrangement (with both parents, a single parent, or without parents). Current suicidal ideation was assessed using one item from the Hopkins Symptom Checklist (SCL-90) [31]. For this analysis, we used the single item on suicide ideation to get restricted and specific information on suicidal ideation in relation to our dependent variables on self-harm and it is found to be a valid approach [32]. Subjects were asked whether, during the previous week, they had had thoughts about ending their life, which was rated on a 1–4 scale (‘‘not at all’’, ‘‘a little’’, ‘‘rather often’’, and ‘‘very often’’). For statistical analysis, this variable was dichotomized into ‘‘none or a little’’ versus ‘‘rather often or very often’’. Substance use variables comprised information about current and past smoking and the frequency of use of various substances (i.e. drinking alcohol to intoxication, use of cannabis, and use of other illicit drugs) in the 12 preceding months. Because of a skew in distribution, the answers were classified into dichotomous categories: ‘‘not smoked’’ versus ‘‘former, occasional, or daily smoker’’; ‘‘not been drunk’’ versus ‘‘been drunk’’; ‘‘not used cannabis’’ versus ‘‘used cannabis’’; and ‘‘not used other illicit drugs’’ versus ‘‘used other illicit drugs’’. Self-perceived health The adolescents were asked how they perceived their current general health status and the response categories were on a five-point ordinal scale ranging from ‘‘very good’’ to ‘‘very poor’’. The distribution on this variable was also very skewed, and the responses were therefore dichotomized into ‘‘good self-perceived health’’ versus ‘‘poor or very poor self-perceived health’’. Depressive symptoms were assessed with six items from the Hopkins Symptom Checklist (SCL-90) [31] using the previous week as a reference period. The items were rated on a 1–4 scale, resulting in a total score ranging from 6 to 24, with higher scores indicating higher levels of depressive symptoms. Current Eating problems were assessed using a Norwegian eight-item version of the Eating Attitudes Test [33, 34]. The items were rated on a 0–3 scale, with total scores ranging from zero to 24. Antisocial behaviour was assessed using 19 variables addressing criminality, rule breaking and other types of antisocial behaviour in the previous 12 months. The variables were derived from a Norwegian version of a questionnaire used originally in the National Youth Longitudinal Study and from the Olweus Scale for Antisocial Behaviour [35]. Those that responded affirmatively were attributed a score of 1 on each of the items and a sum score ranging from 0 to 19 was
123
Author's personal copy 1450
computed, with a higher score reflecting a higher amount of antisocial behaviours in the respondent. Self-esteem was measured using the Global Self-Worth subscale of Harter’s Self-Perception Profile for Adolescents [36], which consists of 10 items scored on a 1–4 scale, with total scores on current self-esteem ranging from 10 to 40, with a higher score indicating higher self-esteem. Loneliness was assessed using the revised UCLA loneliness scale [37], which has been found to have good validity. The 5 items are scored on a 1–4 scale, with total scores on loneliness ranging from 5 to 20, with a higher score indicating the presence of more frequent feelings of loneliness. Statistical analyses Associations between categorical variables were analysed bivariately using cross-tabulation and Chi-squared tests. Differences between group means were tested using the Student’s t test or analysis of variance (ANOVA). In the latter, a subsequent Bonferroni post hoc test was used when the overall test was significant. Effect size was measured using eta squared [38]. Because of the elevated statistical power and an increased risk of obtaining Type I errors, significance was set at P \ 0.01. Stepwise functional discriminant analyses were applied to explore the characteristics of the four categories of self-destructive behaviour and to assess whether these characteristics indicated that the various groups constituted one or several dimensional constructs [39]. Only variables that were significant in bivariate analyses were considered for inclusion in the subsequent discriminant analysis. Wilks’ lambda was used as a selection measure, as it takes into consideration both the differences between the groups and the cohesiveness within the groups [39]. All statistical tests were two-tailed and the statistical analyses were performed using SPSS-15.
Results Prevalence of SH Overall, 9.3 % of the adolescents answered ‘‘yes’’ to the question ‘‘Have you ever taken an overdose of pills or otherwise tried to harm yourself on purpose?’’ and 9.5 % answered ‘‘yes’’ to the question ‘‘Have you ever tried to kill yourself?’’ Nearly half (46.4 %) of the subjects that reported having taken an overdose or otherwise harmed themselves on purpose reported that they had never tried to kill themselves. This group (NSSH) constituted 4.3 % of the whole sample. Nearly half (47.7 %) of the respondents who had tried to kill themselves reported that they had never taken an overdose or otherwise harmed themselves
123
Soc Psychiatry Psychiatr Epidemiol (2013) 48:1447–1455
on purpose. This group (SA) constituted 4.5 % of the overall sample, whereas the prevalence of reporting both suicide attempt and self-harm (SA ? NSSH) was 5 %. SH behaviours related to risk factors Table 1 shows the associations between categories of selfharming behaviour and various sociodemographic variables, current suicidal ideation, substance use, and self-perceived poor health. A significantly greater proportion of girls engaged in any of the reported types of self-harming behaviour compared with boys. The proportion of respondents reporting current suicidal ideation was significantly higher among adolescents in the NSSH ? SA group compared with the other groups; next, the proportion of respondents reporting current suicidal ideation was higher among adolescents in the SA group compared with the NSSH group. As shown in Table 2, the NSSH ? SA group had significantly higher scores on measures of current depressive symptoms, loneliness, antisocial behaviours, and eating problems and lower self-esteem compared with any other group. Subsequent post hoc testing did not alter these observations. Adolescents in the NSSH ? SA group reported self-perceived poorer health more often than did those in any of the other groups. They also reported significantly more substance use on all measures compared with the other groups. Differences between the SH groups Variables on demographics, health, and problem behaviour that were associated bivariately with the SH group were included in a discriminant analysis. The first discriminant function was the only one that was statistically significant and this function explained 94.4 % of the between-group variance. Hence, we considered only this function in the following. In order of relative importance, the following 11 variables contributed to the first discriminant function: frequent suicidal ideation, female gender, self-perceived poor health, smoking, use of illicit drugs, eating problems, having been drunk, use of cannabis, antisocial behaviours, loneliness, and depressive symptoms. F tests showed that these variables discriminated significantly (P \ 0.001) between all pairs of groups. A group centroid (i.e., the mean value of the discriminant function) was calculated for each group and indicated the relative distance between the categories of SH. As shown in Table 3, the largest distance along the canonical discriminant function was found between the categories NSSH ? SA and no SH. The NSSH ? SA group scored highest on the first discriminant function and the SA group
Author's personal copy Soc Psychiatry Psychiatr Epidemiol (2013) 48:1447–1455
1451
Table 1 Association between demographic variables, background characteristics, and history of self-harm No SH, n = 9,083 (%)
NSSH, n = 458 (%)
SA, n = 481 (%)
NSSH ? SA, n = 495 (%)
Statisticsa
48.7
77.2
66.3
71.7***
285.18 (3)
14
24.7
15.9
23.9
18.8
15
32.4
30.3
36.2
33.3
16
36.8
44.8***
32.0
38.6
17
6.2
9.0
7.9
9.3
Both parents
67.9
59.3
51.4
54.5
Mother or father Others
32.1 3.7
40.7 4.2
48.6*** 5.2
45.5 11.5***
175.38 (6)
3.6
14.7
34.4
55.5***
346.48 (3)
Variable Sociodemographic variables Gender Female Age (years)
46.70 (9)
Living with
Current suicidal ideationb Substance use Smoking
25.0
51.8
49.6
66.5***
940.38 (9)
Been drunk
37.3
58.7
50.9
65.6***
269.15 (3)
Cannabis
6.6
19.5
14.7
31.3***
488.67 (3)
Other
3.5
9.3
8.7
23.6***
471.92 (3)
2.2
6.4
8.0
20.1***
528.39 (3)
Use of narcotics
Health Self-perceived poor health a
Chi-squared tests were used for categorical variables
b
Categories ‘‘quite often to very often’’
Table 2 Mean psychological measures among adolescents not involved in SH and individuals in the SH groups Variable
No SH (n = 9,083)
NSSH (n = 458)
SA (n = 481)
NSSH ? SA (n = 495)
Effect size (%)a
Statisticsb
Depressive symptoms
11.3 (4.0)
15.3 (4.3)
15.4 (4.7)
17.7 (4.7)***
14.7
600.6 (3)
Eating problems
5.6 (4.7)
8.8 (5.2)
8.6 (5.7)
10.3 (6.1)***
6.5
234.4 (3)
Antisocial behaviour
2.7 (3.2)
4.1 (3.9)
4.2 (4.2)
5.9 (5.2)***
4.8
185.0 (3)
32.7 (5.0) 8.1 (2.6)
29.2 (5.4) 9.3 (2.9)
29.3 (6.3) 9.7 (3.6)
26.9 (6.7)*** 10.5 (3.6)***
8.1 5.1
299.0 (3) 177.0 (3)
Self-esteemc Loneliness
SD is indicated within brackets *** P \ 0.001 for all post hoc comparisons: NSSH ? SA [ SA = NSSH [ no SH a
Eta squared
b
Results of one-way ANOVA with F values (df)
c
A high score indicates better self-esteem
had the second-highest score, which was somewhat higher than that of the NSSH only group.
Discussion The most important finding of this study, which collected data on both nonsuicidal self-harm and suicide attempts in a large and representative sample of adolescents, was that one-third of adolescents with a history of SH reported both
NSSH and SA. These adolescents were significantly more likely to report current suicidal ideation and a range of other health and behavioural problems. Few previous studies have focused on this subgroup of self-harming adolescents and those who have, are either based on clinical samples [19, 23, 40, 41] or they have used relatively small community samples [22, 26]. In our study, an underlying dimension of mental ill-health and behavioural problems was able to discriminate between groups of adolescents reporting different types of SH. This dimension
123
Author's personal copy 1452
Soc Psychiatry Psychiatr Epidemiol (2013) 48:1447–1455
Table 3 Results of the discriminant analysis Variable
Standardized canonical coefficients of discriminant function 1
Suicidal ideation
0.778
Gender
0.187
Self-perceived poor health Smoking
0.151 0.144
Illicit-drug use
0.117
Eating problems
0.111
Been drunk
0.102
Smoked cannabis
0.077
Antisocial behaviour
0.069
Loneliness
0.069
Depressive symptoms
0.059
Group centroids
No SH
-0.279
NSSH
0.955
SA
1.616
NSSH ? SA
2.812
could be important to take into consideration in future studies of SH in adolescents. The lifetime prevalence rate of self-harming behaviour (self-poisoning or self-injury irrespective of the intent) found in this study (9.3 %) corresponds well with findings from previous community-sample studies of adolescents (13.2 %) [11, 27, 42]. The further sub-categorization of SH into three groups has not been done in large-scale population sample studies, but it resembles that of clinical studies, yet the prevalence rates of these categories are not comparable across clinical and population samples. Two recent studies conducted in Norway, not comparable with the present study, should be mentioned. None of them addressed the NSSH ? SA group, but its worth notifying that the first study found percentages of 4.9 % for engaging in NSSH only and 8.7 % for SA only, and the last found percentages of 2.9 % in the NSSH only group, whereas 3.0 % in the SA group [25, 27]. In addition, the demographic, mental health, and behavioural characteristics of the self-harming adolescents in this study were also in accordance with findings from earlier epidemiological studies [10]. The significant co-occurrence of both NSSH and SA has been reported in some previous studies [22]. Our finding that adolescents exhibiting both NSSH and SA reported more depressive symptoms and suicidal ideation compared with those who reported NSSH only is in line with the findings of previous communitysample studies [22, 26, 43], as well as with those of clinicalsample studies [23, 44, 45]. The occurrence of more severe mental health problems and behavioural problems beyond covariates that are specific to SH groups also suggest that our
123
findings are in accordance with those reported in previous studies. Even though studying only suicide attempters, it is interesting that a recent epidemiological study of selfreported suicide attempts identified three classes of attempters who were distinguishable according to their levels of substance use and violent behaviours. Depression was high among all three groups [46]. Research on the co-occurrence of NSSH and SA remains limited [44] and the nature of the relationship between these two behaviours is ambiguous [47]. First, their temporal relationship is rarely addressed [25], despite the fact that it has been suggested that NSSH serves as a harbinger of distress that may lead some individuals to attempt suicide [47]. Moreover, research on unique and overlapping factors regarding NSSH and SA has been limited regarding the investigation of underlying shared aetiological factors between the forms [22]. Thus, it is not clear why adolescents exhibiting both NSSH and SA were more burdened with mental health and behavioural problems compared with other self-harmers, however, it is possible that a higher distress burden over time and an increase in the number of critical events may lead to both NSSH and SA. It is also possible that hopelessness is a key factor in distinguishing those who perform both NSSH and SA from those exhibiting only NSSH, given its close association with depression and negative self-evaluation and the fact that greater levels of the two latter manifestations have been found in those who experience both NSSH and SA [17, 18, 20]. Acknowledging the research and clinical impression of the co-occurrence of the behaviours, researchers have begun to investigate risk factors specific to NSSH or attempted suicide. Some researchers argue that various forms of SH are the same phenomena, but represent different degrees of suicidality, whereas others claim that it is different phenomena. The last view is articulated in a proposal for DSM-V to identify nonsuicidal self-injury as a separate clinical disorder, even though empirical knowledge about the prevalence and course of NSSH to date is scarce. The absence of intent and the emotion regulation motives for NSSH are understood as coping strategies to endure life. Regarding SA, the intent and motives often circle around getting rid of emotions and ending life [48]. This can be argued, are so different intent and motives, that it could support the view that they are separate phenomena. On the other side, the results of some studies suggest that co-morbid engagement in NSSH with other problem behaviours, such as suicide attempts is so salient [22, 26], which could support the view of a continuum hypothesis. By directly comparing risk factors in both attempted suicide and NSSH, researchers will be able to establish the factors that differentiate NSSH from attempted suicide, as well as the factors common to both behaviours. Our
Author's personal copy Soc Psychiatry Psychiatr Epidemiol (2013) 48:1447–1455
findings suggests that there are various degrees of suicidality among those involved, but our study cannot contribute in particular to the debate on whether nonsuicidal self-injury should be a separate clinical syndrome or not. Strengths and limitations The present study has several strengths. First, few previous studies based on community samples distinguished between different categories and combinations of selfharming behaviour [44]. Here, we achieved this distinction using two different questions on self-harming behaviour. Furthermore, our study sample was much larger than any of the previous community-sample studies performed in this area and comprised a broad array of relevant psychosocial measures; therefore, it allowed the exploration of various possible and complex associations in multivariate analyses. A large population sample size with high response rate and hence representativeness, is in contrast to what has typified earlier prevalence studies with convenience samples or small clinical samples. Compared to previous clinicalsample studies, it is likely that adolescents who self-harm by cutting or other forms of self-injuries are better represented in this study. Research has shown that those who self-harm by overdose are more likely to seek help and are admitted to hospital care more often than those who cut or injure themselves in other ways. In previous research about 80 % of people who go to acute and emergency departments following an act of self-harm have taken an overdose, even though behaviours such as cutting may be twice as prevalent as overdose in those who self-harm in the community [49]. However, some study limitations should also be noted. First, a cross-sectional study design such as ours does not allow causal inferences. Second, conducting a large-scale survey with such a large sample does not allow for in-depth questions on all topics, and our survey comprised only two single questions on self-harming behaviour. However, data from large school surveys in the CASE study [8, 50] showed that among adolescents who reported SH, a large majority were also classified as SH on the basis of expert evaluations of detailed reports on the actual behaviour. In the Norwegian part of the CASE study 77 % of the respondents reporting life time SH were classified as SH by the expert evaluation [51]. This would suggest that the prevalence of SH may be somewhat over-estimated, when assessed with this single question. On the other hand, while the definition of self-harm covers all methods, medication overdose is specifically mentioned in the SH question. If this implies that self-harmers who have applied other, and more common self-harming methods, such as cutting and burning, are less likely to report this as SH, the prevalence of SH will be under-estimated. Consequently, there are
1453
various factors that may lead to both an upward bias and a downward bias of the prevalence estimate when applying a single question about SH as was done in this study and in many previous studies. Third, we lacked information on the temporal relationship between NSSH, SA and the other psychosocial problems. The two questions on self-harming behaviour referred to lifetime experiences, whereas the various psychosocial characteristics assessed pertained to experiences of the past year or of the past week, which may have deflated the associations between categories of selfharming behaviour and psychosocial variables. Fourth, a number of variables that are previously found to be important correlates to SH were not included in this study. If measures on other known risk factors for SH had been available, such as impulsivity, emotion regulation, temperament, and family relations [2, 23], we could have obtained an even more detailed description of characteristics of the various SH groups. Implications for clinical and preventive practice and further research Our results suggest that adolescents alternating between NSSH and SA behaviours are particularly burdened with mental ill-health and behavioural problems. Thus, this group may constitute a particular challenge for clinicians and school health personnel regarding assessment and provision of adequate help and services. Adolescents at risk should be identified more effectively by asking about SH in cases presenting to health services for various psychosocial problems. Future surveys and health care providers that are in contact with adolescents at risk for SH should assess current and past SH by enquiring about both NSSH and SA behaviours. Moreover, clinicians should be aware that many adolescents perform both types of behaviour and it is important that they avoid clinical stereotypes regarding suicidal intent in adolescents with self-harming behaviours [19, 52]. Given the particular burden of distress that adolescents with co-occurring experiences of NSSH and SA seem to experience, the challenges of further research in this area appear particularly important with respect to this group and are of several types. First, the nature and the temporal sequencing of co-occurring experiences of NSSH and SA are rarely addressed and are not well understood [11]. To date, we know little about the shared and nonshared features of NSSH and SA among adolescents; thus, further research on the characteristics and aetiology of these behaviours is needed [40]. However, it has been proposed that NSSH and other potentially harmful behaviours serve similar functions; therefore, careful examination of their cooccurrence is a necessary step in the testing of this conceptualization of SH [53]. Regarding temporal sequencing,
123
Author's personal copy 1454
Joiner and co-workers hypothesized that NSSH is often a starting point, as suicide is such a frightening and extreme action and most people initially lack the ability to engage in suicide attempts; however, after engaging in repeated NSSH, individuals may become more courageous, competent, and willing to make suicide attempts [40, 41, 54]. To date, few studies have addressed this empirically and the results are discrepant [25]. Thus, the need for further research on the temporal association between NSSH and SA mentioned by Jacobson and Gould [11] still applies. Moreover, it seems particularly important to assess whether adolescents that alternate between NSSH and SA being particularly vulnerable and in need of help services receive such services, and whether specific treatment programmes and care services are effective in helping these adolescents with multiple problems. Acknowledgments The assistance of John Eriksen, PhD, and of the Norwegian Social Research (NOVA) is gratefully acknowledged. Finance was supported by the Norwegian Extra Foundation for Health and Rehabilitation through EXTRA funds and from the National Centre for Suicide Research and Prevention, University of Oslo, Norway. The authors declare that they have no conflict of interest.
References 1. Cooper J, Kapur N, Webb R, Lawlor M, Guthrie E, MackwayJones K, Appleby L (2005) Suicide after deliberate self-harm: a 4-year cohort study. Am J Psychiatry 162:297–303 2. Gould MS, Greenberg T, Velting DM, Shaffer D (2003) Youth suicide risk and preventive interventions: a review of the past 10 years. J Am Acad Child Adolesc Psychiatry 42:386–405 3. Hall DJ, O’Brien F, Stark C, Pelosi A, Smith H (1998) Thirteenyear follow-up of deliberate self-harm, using linked data. Br J Psychiatry 172:239–242 4. Groholt B, Ekeberg O, Wichstrom L, Haldorsen T (1997) Youth suicide in Norway, 1990–1992: a comparison between children and adolescents completing suicide and age- and gender-matched controls. Suicide Life Threat Behav 27:250–263 5. Hawton K, Zahl D, Weatherall R (2003) Suicide following deliberate self-harm: long-term follow-up of patients who presented to a general hospital. Br J Psychiatry 182:537–542 6. Lewinsohn PM, Rohde P, Seeley JR (1996) Adolescent suicidal ideation and attempts: prevalence, risk factors, and clinical implications. Clin Psychol 3:25–46 7. Nock MK, Borges G, Bromet EJ et al (2008) Cross-national prevalence and risk factors for suicidal ideation, plans and attempts. Br J Psychiatry 192:98–105 8. Hawton K, Rodham K, Evans E, Weatherall R (2002) Deliberate self harm in adolescents: self report survey in schools in England. BMJ 325:1207–1211 9. Bridge JA, Goldstein TR, Brent DA (2006) Adolescent suicide and suicidal behavior. J Child Psychol Psychiatry 47:372–394 10. Evans E, Hawton K, Rodham K (2004) Factors associated with suicidal phenomena in adolescents: a systematic review of population-based studies. Clin Psychol Rev 24:957–979 11. Jacobson CM, Gould M (2007) The epidemiology and phenomenology of non-suicidal self-injurious behavior among adolescents: a critical review of the literature. Arch Suicide Res 11:129–147
123
Soc Psychiatry Psychiatr Epidemiol (2013) 48:1447–1455 12. Haw C, Hawton K (2011) Problem drug use, drug misuse and deliberate self-harm: trends and patient characteristics, with a focus on young people, Oxford, 1993–2006. Soc Psychiatry Psychiatr Epidemiol 46:85–93 13. Skegg K (2005) Self-harm. Lancet 366:1471–1483 14. Hawton K, Bergen H, Casey D et al (2007) Self-harm in England: a tale of three cities. Soc Psychiatry Psychiatr Epidemiol 42: 513–521 15. Schmidtke A, Bille-Brahe U, DeLeo D et al (1996) Attempted suicide in Europe: rates, trends and sociodemographic characteristics of suicide attempters during the period 1989–1992. Results of the WHO/EURO Multicentre Study on Parasuicide. Acta Psychiatr Scand 93:327–338 16. Michel K, Ballinari P, Bille-Brahe U et al (2000) Methods used for parasuicide: results of the WHO/EURO Multicentre Study on Parasuicide. Soc Psychiatry Psychiatr Epidemiol 35:156–163 17. Nixon MK, Cloutier P, Jansson SMP (2008) Nonsuicidal selfharm in youth: a population-based survey. CMAJ 178:306–312 18. De Leo D, Heller TS (2004) Who are the kids who self-harm? An Australian self-report school survey. Med J Aust 181:140–144 19. Stanley B, Gameroff MJ, Michalsen V, Mann JJ (2001) Are suicide attempters who self-mutilate a unique population? Am J Psychiatry 158:427–432 20. Shaffer D, Jacobson CM (2011) Proposal to the DSM-V childhood disorder and mood disorder work groups to include nonsuicidal self-injury (NSSI) as a DSM-V disorder. American Psychiatric Association. http://www.dsm5.org/Proposed%20 Revision%20Attachments/APA%20DSM-5%20NSSI%20Proposal. pdf. Accessed 15 January 2011 21. Muehlenkamp JJ (2005) Self-injurious behavior as a separate clinical syndrome. Am J Orthopsychiatry 75:324–333 22. Muehlenkamp JJ, Gutierrez PM (2007) Risk for suicide attempts among adolescents who engage in non-suicidal self-injury. Arch Suicide Res 11:69–82 23. Jacobson CM, Muehlenkamp JJ, Miller AL, Turner JB (2008) Psychiatric impairment among adolescents engaging in different types of deliberate self-harm. J Clin Child Adolesc Psychol 37:363–375 24. Guertin T, Lloyd-Richardson E, Spirito A, Donaldson D, Boergers J (2001) Self-mutilative behavior in adolescents who attempt suicide by overdose. J Am Acad Child Adolesc Psychiatry 40:1062–1069 25. Wichstrom L (2009) Predictors of non-suicidal self-injury versus attempted suicide: similar or different? Arch Suicide Res 13: 105–122 26. Brausch AM, Gutierrez PM (2010) Differences in non-suicidal self-injury and suicide attempts in adolescents. J Youth Adolesc 39:233–242 27. Larsson B, Sund AM (2008) Prevalence, course, incidence, and 1-year prediction of deliberate self-harm and suicide attempts in early Norwegian school adolescents. Suicide Life Threat Behav 38:152–165 28. Groholt B, Ekeberg O, Wichstrom L, Haldorsen T (2000) Young suicide attempters: a comparison between a clinical and an epidemiological sample. J Am Acad Child Adolesc Psychiatry 39: 868–875 29. Madge N, Hewitt A, Hawton K, Wilde EJ, Corcoran P, Fekete S, Heeringen KV, Leo DD, Ystgaard M (2008) Deliberate self-harm within an international community sample of young people: comparative findings from the Child and Adolescent Self-harm in Europe (CASE) Study. J Child Psychol Psychiatry 49(6):667–677 30. Wichstrom L (2000) Predictors of adolescent suicide attempts: a nationally representative longitudinal study of Norwegian adolescents. J Am Acad Child Adolesc Psychiatry 39:603–610
Author's personal copy Soc Psychiatry Psychiatr Epidemiol (2013) 48:1447–1455 31. Derogatis LR, Lipman RS, Rickels K, Uhlenhuth EH, Covi L (1974) The Hopkins Symptom Checklist (HSCL): a self-report symptom inventory. Behav Sci 19:1–15 32. Desseilles M, Perroud N, Guillaume S, Jaussent I, Genty C, Malafosse A, Courtet P (2012) Is it valid to measure suicidal ideation by depression rating scales? J Affect Disord 136:398– 404 33. Lavik NJ, Clausen SE, Pedersen W (1991) Eating behaviour, drug use, psychopathology and parental bonding in adolescents in Norway. Acta Psychiatr Scand 84:387–390 34. Garner DM, Olmsted MP, Bohr Y, Garfinkel PE (1982) The eating attitudes test: psychometric features and clinical correlates. Psychol Med 12:871–878 35. Windle M (1990) A longitudinal study of antisocial behaviors in early adolescence as predictors of late adolescent substance use: gender and ethnic group differences. J Abnorm Psychol 99:86–91 36. Harter S (1988) Manual for the self-perception profile for adolescents. University of Denver, Denver 37. Russell D, Peplau LA, Cutrona CE (1980) The revised UCLA Loneliness Scale: concurrent and discriminant validity evidence. J Pers Soc Psychol 39:472–480 38. Cohen J (1988) Statistical power analysis for the behavioral sciences. Erlbaum Associates, Hillsdale 39. Klecka WR (1980) Discriminant analysis. Sage, Beverly Hills 40. Boxer P (2010) Variations in risk and treatment factors among adolescents engaging in different types of deliberate self-harm in an inpatient sample. J Clin Child Adolesc Psychol 39:470–480 41. Nock MK, Joiner TE Jr, Gordon KH, Lloyd-Richardson E, Prinstein MJ (2006) Non-suicidal self-injury among adolescents: diagnostic correlates and relation to suicide attempts. Psychiatry Res 144:65–72 42. Evans E, Hawton K, Rodham K, Deeks J (2005) The prevalence of suicidal phenomena in adolescents: a systematic review of population-based studies. Suicide Life Threat Behav 35:239–250 43. Wong JP, Stewart SM, Ho SY, Lam TH (2007) Risk factors associated with suicide attempts and other self-injury among Hong Kong adolescents. Suicide Life Threat Behav 37:453–466
1455 44. Cloutier P, Martin J, Kennedy A, Nixon MK, Muehlenkamp JJ (2010) Characteristics and co-occurrence of adolescent non-suicidal self-injury and suicidal behaviours in pediatric emergency crisis services. J Youth Adolesc 39:259–269 45. Dougherty DM, Mathias CW, Marsh-Richard DM, Prevette KN, Dawes MA, Hatzis ES, Palmes G, Nouvion SO (2009) Impulsivity and clinical symptoms among adolescents with non-suicidal self-injury with or without attempted suicide. Psychiatry Res 169:22–27 46. Pena J, Matthieu M, Zayas L, Masyn K, Caine E (2012) Cooccurring risk behaviors among White, Black, and Hispanic US high school adolescents with suicide attempts requiring medical attention, 1999–2007: implications for future prevention initiatives. Soc Psychiatry Psychiatr Epidemiol 47:29–42 47. Whitlock J (2010) Self-injurious behavior in adolescents. PLoS Med 7:e1000240 48. Muehlenkamp JJ, Gutierrez PM (2004) An investigation of differences between self-injurious behavior and suicide attempts in a sample of adolescents. Suicide Life Threat Behav 34:12–23 49. Horrocks J, Price S, House A, Owens D (2003) Self-injury attendances in the accident and emergency department: clinical database study. Br J Psychiatry 183:34–39 50. Hawton K, Rodham K, Evans E (2006) By their own young hand. Deliberate self-harm and suicidal ideas. Jessica Kingsley, London 51. Ystgaard M, Reinholdt NP, Husby J, Mehlum L (2003) Deliberate self harm in adolescents. [Norwegian]. Tidsskr Nor Laegeforen 123:2241–2245 52. Groholt B, Ekeberg O, Haldorsen T (2000) Adolescents hospitalised with deliberate self-harm: the significance of an intention to die. Eur Child Adolesc Psychiatry 9:244–254 53. Nock MK (2009) Why do people hurt themselves? New insights into the nature and functions of self-injury. Curr Dir Psychol Sci 18:78–83 54. Joiner TE Jr, Brown JS, Wingate LR (2005) The psychology and neurobiology of suicidal behavior. Annu Rev Psychol 56:287– 314
123
