Research on adolescent development has devoted efforts to the understanding of the roots of potentially self damaging behaviors, including suicide, eating disorders, substance abuse, sexual promiscuity, risk-taking, violence and aggression, delinquency and, more recently, self-harm or non-suicidal self-injurious behavior (NSSI). Literature data have shown that adolescence is a critical period for the onset of self-harm . The prevalence of NSSI among adolescents in community based studies range between 13% and 28% [2–4]; in general, literature suggests a seemingly increasing prevalence of such behaviors in the teenager population [5, 6].
Not surprisingly, higher rates of self-harm are apparent in individuals receiving mental health treatment: NSSI occurs in about 20% of adult psychiatric patients  and in 40-80% of adolescent psychiatric patients [8–10].
NSSI has also been described as one of the most diffuse and challenging clinical phenomena reported in adolescent inpatient samples [11, 12]. A recent review study of discharge diagnoses indicated a threefold increase in NSSI among hospitalized adolescents from 1990 to 2000 .
Much debate concerns the psychopathological meaning of NSSI, due to its uncertain boundaries and heterogeneous manifestations. Overall, literature on psychopathological characteristics of NSSI in adolescence has considered separately two distinct controversial aspects: the relationship between NSSI and personality disorders, and the relationship between NSSI and the depressive-suicidal dimension. Moreover, adolescents with severe personality disorders are often assessed and treated in different clinical settings and with different approaches than those with mood disorders and suicidality.
Empirical findings show that the psychopathological dimension more consistently related to NSSI concerns personality functioning. In particular, a close link has been evidenced between NSSI and Borderline Personality Disorder (BPD) [14–17]. The majority of studies concern adult population. Indeed, epidemiological data show that 80% of adult BPD patients have exhibited at least one episode of self-harm . In DSM-IV , self-harm has been represented under criterion 5 of BPD: keeping in mind the controversies about diagnosing personality disorders prior to adulthood and the fact that considering self-harm pathognomonic of BPD could lead to inappropriate management, the "DSM-5 Childhood and Adolescent Work Group" is now recommending the inclusion of the new diagnosis: Non-suicidal Self-Injury (NSSI). The proposed new diagnosis of NSSI applies to individuals engaging in intentional self-inflicted damage on 5 or more days in the last year, without suicidal intent and presenting a significant distress or impairment. The inclusion of the new diagnosis may reflect the clear cut relation between NSSI and childhood/adolescence, reduce the automatic assumption that an adolescent who engages in NSSI may have BPD and hopefully promote research to further clinical guidelines for treatment [20, 21]. Since NSSI is a distinct aspect from BPD, it is important to fully articulate the relationship between NSSI and personality functioning in adolescence. The analysis of personality features of adolescent inpatient and outpatient populations exhibiting NSSI have so far confirmed the typical adult association between NSSI and BPD [21–23].
Substantial research attention has been given to the presence of specific forms of psychopathology associated with NSSI among hospitalized adolescent: it has been suggested  that most adolescent inpatients engaging in NSSI meet criteria for a DSM-IV Axis I diagnosis, with elevated rates of Major Depressive Disorder (MDD) (42%), Post-Traumatic Stress Disorder (PTSD) (24%), Substance Use Disorder (SUD) (60%). In a retrospective chart review using medical records, Jacobson et al.  found 67% of MDD in the total sample examined (NSSI and NSSI "plus suicide attempts" outpatients); Muehlenkamp et al. , using a similar methodology, examined how BPD symptoms relate to suicide attempts or NSSI within a population of adolescent outpatients, finding two BPD features ("confusion about self" and "unstable interpersonal relationships") as distinct predictors of "NSSI" and "NSSI + Suicide" group status, but not a strong variation in the impact of the single features on the different subgroups.
Although the inclusion of the new diagnosis makes clear the intention to consider NSSI and suicide attempts as distinct phenomena, several important questions are yet to be explained. First, NSSI and suicide attempts could co-occur with different modalities in different clinical populations: a sizable portion of self-injurers (50% of outpatients; 70% of inpatients) reports having attempted suicide at least once [5–22]. Some epidemiological and research data evidence that many suicides are not preceded by NSSI. In general population samples NSSI seems to have less severe consequences than attempted suicide and a different risk trajectory [5, 6], but in adolescent inpatients who have attempted suicide a history of NSSI before the index episode is more likely than in those who have only suicidal ideation . Data from ADAPT study show that in depressed adolescents receiving treatment over a 6 months follow-up, NSSI at baseline is an independent predictor of suicide attempt, even stronger than a history of suicide attempt itself ; Asarnow and colleagues identified similar findings in adolescents with treatment- resistant depression .
Second, although prior research has focused on the identifications of possible psychopathological links between NSSI, depression and suicidal ideation this crucial question remains unclear. Depressive symptoms seem to distinguish "NSSI-only" patients from NSSI patients who attempt suicide , thus implying a role for depressive conditions in the escalation from NSSI to suicidal behaviors. Moreover, self-harm has been found to be associated with depressive ideation, including feeling repulsed by life, having greater amounts of apathy, self-blame, and fewer connections to family members . By definition, explicit suicidal ideation would not pertain NSSI subjects who, nonetheless, may end up in attempting suicide. Therefore, the individuation of instruments aimed in helping clinicians identify a suicide risk when there is not an explicit suicidal ideation seems a necessary step. Ideational factors related to feelings toward life and death has proved a useful construct in discriminating suicidal adolescents, non-suicidal adolescents and a psychiatric control group .
Given these considerations in this study we pursue two main objectives:
1) To describe the characteristics of NSSI and related psychopathology/personality functioning in a sample of NSSI adolescent inpatients;
2) To investigate whether characteristics such as depression symptoms and attitude toward life and death discriminate between NSSI subjects who have attempted suicide (NSSI - SA) and NSSI subjects who have not attempted suicide (NSSI only).