An examination of nonsuicidal self-injury among college students.

This study examines characteristics (i.e., prevalence, method, age of onset, frequency) of nonsuicidal self-injury (NSSI) and associated risk factors in a college student sample. Results revealed 11.68% admitted to engaging in NSSI at least once and no significant gender difference in occurrence of NSSI. Even in this college sample, those who self-injure differed substantially from non-self-injurers with regard to emotion regulation, but were not found to differ significantly on either early attachment or childhood trauma and abuse. Importance of understanding NSSI as an emerging behavior among college students is discussed.

**********

The occurrence of nonsuicidal self-injury (NSSI) amongst college students has increasingly come to the attention of the mental health professionals (Whitlock, Purington, Eells, & Cummings, 2006). In a recent survey of school counselors' experiences with NSSI, the major concern expressed was a lack of training in this area and the need to be informed about the occurrence and characteristics of NSSI (Heath, Toste, & Beettam, 2007). The present paper provides information on the basic facts and associated risk factors of NSSI in college students.

Defining the exact parameters of NSSI behavior has not been straightforward and interpreting the research in the field can be challenging due to differences in the operationalization of the definition. In response to these concerns, the International Network for the Study of Self-injury (ISSS) was established in 2006 by leading researchers in the field of self-injury to work towards a consensus regarding key issues. One year later, in June 2007, the ISSS agreed on the following definition of NSSI:

   The deliberate, self-inflicted destruction of body tissue resulting
   in immediate damage, without suicidal intent and for purposes not
   socially sanctioned. As such, this behavior is distinguished from:
   suicidal behaviors involving an intent to die, drug overdoses, and
   other forms of self-injurious behaviors, including
   culturally-sanctioned behaviors performed for display or aesthetic
   purposes; repetitive, stereotypical forms found among individuals
   with developmental disorders and cognitive disabilities, and severe
   forms (e.g., self-immolation and auto-castration) found among
   individuals with psychosis. (ISSS, 2007)

However, NSSI can be understood as a subset of the larger range of self-harming behaviors. Deliberate self-harm, as defined by the Child and Adolescent Self-harm (CASE) group in Europe, is an act with a nonfatal outcome in which an individual deliberately does one or more of the following: initiated behavior (e.g., self-cutting, jumping from a height) intended to cause self-harm; ingested a substance in excess of the prescribed or generally recognized therapeutic dose; ingested a recreational or illicit drug that was an act that the person regarded as self-harm; or ingested a non-ingestible substance or object, irrespective of suicidal intent (Hawton, Rodham, Evans, & Weatherall, 2002; Hawton et al., 2003). More recently, Hawton and colleagues have sought to change the term from deliberate self-harm to self-harm (Hawton & James, 2005).

Clearly, self-harm is a broader construct than NSSI and critical readers of the literature should be aware that NSSI, while subsumed under the self-harm definition, cannot be equated with other self-harming behaviors. In their classification of suicide-related behaviors, Silverman et al. (2007) categorize each behavior on the basis of suicidal intent (e.g., none, undetermined, some) and outcome (e.g., fatal injury, non-fatal injury, no injury). Silverman and colleagues emphasize the importance of determining the suicidal intent or motivation behind the behavior. Thus, while in the past deliberate self-harm as studied in Europe has largely not evaluated suicidal intent, Silverman et al. insist that in the future, behaviors that differ in suicidal intent cannot be equated. In summary, while NSSI may be related to other suicidal behaviors, it is a distinct and separate behavior from either suicide attempts or the broader deliberate self-harm definition. When one reviews the literature, or is attempting to identify NSSI in a clinical setting, it is essential to distinguish between the range of self-harm behaviors that may have suicidal intent and NSSI, which does not.

Many assertions have been made about the increase of "self-injury" in youth and young adults (e.g., Adler & Adler, 2007; Favazza, 1998) and this behavior has been referred to as the new youth epidemic or the "new" anorexia (Shaw, 2002; Zila & Kiselica, 2001). However, the empirical evidence documenting an increase in self-injury has been limited to studies of the broader construct of deliberate self-harm (e.g., Hawton, Fagg, Simkin, Bale, & Bond, 2000; Hawton et al., 2003). Although no empirical research presents results directly demonstrating an increase, Whitlock et al. (2006) found that counselors in a college setting reported that an increased number of clients were coming forward with NSSI. Similarly, Heath, Toste, and Beettam (2006) found that high school teachers perceived NSSI behavior to be on the rise. Thus, although it appears that NSSI is increasing, this could be a result of students becoming more willing to disclose the behavior and seek support. In the following section, a brief review of studies examining the prevalence of NSSI in youth and young adults will be presented and research examining the risk factors for NSSI in community samples will be discussed.

Prevalence of Self-Injury in the Community

Researchers have investigated the prevalence of NSSI in community samples of both adolescents and young adults. In general, lifetime prevalence rates for adolescents and young adults in the community range between 10% to 20% (e.g., Muehlenkamp & Gutierrez, 2007; Ross & Heath, 2002; Whitlock, Eckenrode, & Silverman, 2006). Where notably discrepant results are found (e.g., Gratz, 2006; Lloyd-Richardson, Perrine, Dierker, & Kelley, 2007) with much higher prevalence rates, checklists of all possible self-injurious behaviors were presented to the participants and they were asked to indicate every behavior in which they had engaged. This method consistently results in higher incidence reporting. By providing a list of many possible forms of tissue damage (e.g., biting, sticking pins in skin, scraping skin, interfering with wound healing), the definition is effectively broadened as many participants would not think to identify themselves as a "self-injurer" on the basis of interference with wound healing, one of the listed behaviors. In addition, participants' understanding of the behavior may not be consistent with the researchers' conceptualization. For example, "sticking pins into skin" may be a NSSI behavior or it may be related to drug use, self-tattooing, or other body modifications. Ideally, with a checklist, a follow-up interview should be conducted to clarify the occurrence and intent of the behavior.

Another notable finding in reviewing the literature on prevalence of NSSI is that there is more likely to be a gender difference when examining clinical samples than community samples (e.g., Jacobsen, Muehlenkamp, & Miller, 2006; Nixon, Cloutier, & Aggarwal, 2002). This may be because women are more prone to seek help than men, or due to the inclusion of overdose or inappropriate ingestion of medication without suicidal intent in many of the studies conducted in clinical settings, which have been found to be largely female behaviors (e.g., Briere & Gil, 1998; Rodham, Hawton, & Evans, 2004). In examining the prevalence of NSSI for male and female participants within community samples, it appears that the inclusion of overdose and ingestion of substances may be at the root of the observed gender differences. Specifically, studies of self-injury in community samples that have revealed gender differences (e.g., Laye-Gindhu & Schonert-Reichl, 2005; Nixon, Cloutier, & Jansson, 2007; Patton et al., 1997) have included overdose or abuse of pills/medication without suicide intent. Studies that have limited their definition to behaviors such as cutting, burning, self-hitting, and other forms of direct tissue damage have failed to find gender differences (e.g., Izutsu et al., 2006; Lloyd-Richardson et al., 2007; Muehlenkamp & Gutierrez, 2004; 2007; Ross & Heath, 2002; Zoroglu et al., 2003). This pattern holds true in studies exploring NSSI among young adults; with the exception of Whitlock, Eckenrode, and Silverman (2006), who reported a very small gender difference. Thus, NSSI may not necessarily be a predominately female behavior, although in clinical samples there are significantly more women who engage in NSSI than men.

Risk Factors for Nonsuicidal Self-Injury

Gratz (2006) noted that risk factors for NSSI may be understood as falling into one of two general categories; environmental risk factors (e.g., childhood maltreatment) or individual risk factors (e.g., difficulties with emotion expression and intensity). It is asserted that, historically, the majority of the self-injury literature has focused on the environmental risk factors. As such, researchers have consistently found a link between childhood sexual abuse and adult engagement in a variety of self-harming behaviors (e.g., Boudewyn & Liem, 1995), including NSSI (e.g., Gratz, Conrad, & Roemer, 2002; Zoroglu et al., 2003). Physical abuse in childhood has also been suggested as a possible risk factor for later self-harming behavior (e.g., Carroll, Schaffer, Spensley, & Abamowitz, 1980; Gratz et al., 2002). However, very little of the research has drawn on more normative community samples and those that have (e.g., Gratz, 2006) have had samples that may suffer from significant selection bias. In addition, clinicians and theoreticians have suggested that early attachment difficulties or disruption may contribute to later self-harm (e.g., Conterio & Lader, 1998; Suyemoto, 1998; van der Kolk, 1996; Walsh, 2006). The study conducted by Gratz and colleagues (2002) is one of the only studies examining the role of the parent-child relationship as a risk factor for NSSI. Results revealed that emotional neglect and the quality of the parent-child bond were both associated with risk for NSSI in adulthood. These findings clearly point to the importance of investigating both childhood trauma (physical or sexual abuse) and the quality of parent-child attachment when studying the risk factors for NSSI in community samples.

Interestingly, the occurrence of childhood maltreatment or trauma is believed to result in individuals developing later difficulties with emotion regulation (e.g., Shields & Cicchetti, 1998; van der Kolk & Fisher, 1994), which in turn has been suggested to be a risk factor for NSSI (Gratz et al., 2002; Klonsky, 2007). The association between emotion dysregulation and NSSI has been reported in studies of college undergraduates (Gratz & Roemer, 2004) and stratified community samples of adults (Briere & Gil, 1998), as well as high-risk groups such as self-injuring male prisoners (Haines, Williams, Brain, & Wilson, 1995). However, the results of the reviewed studies are limited by the samples and the design. In effect, most participants come either from high-risk environments (Ayton, Rasool, & Cottrell, 2003; Tulloch, Blizzard, & Pinkus, 1997) or represent a self-selected sample of individuals willing to participate in a study of NSSI (Gratz et al., 2002). Moreover, there is often no control group that would allow for a more exact comparison. Nevertheless, there is support for the idea that individuals in the community who engage in NSSI do so to obtain relief from intense emotions (Briere & Gil, 1998; D'Onofrio, 2007; Gratz & Roemer, 2004). Specifically, individuals seem to employ NSSI as a way to regulate overwhelming emotions such as hostility and anxiety, although it seems to represent a very short-term relief (Favazza & Conterio, 1989; Ross & Heath, 2002; Walsh, 2006).

Thus, in examining the literature, there has been theoretical, clinical, and empirical support for the hypotheses that early childhood abuse and attachment difficulties serve as environmental risk factors for the development of NSSI behaviors and that emotion dysregulation, possibly caused by these environmental stressors, is a consistent individual risk factor. However, only Gratz and colleagues (Gratz, 2006; Gratz et al., 2002) have attempted to simultaneously study both environmental and individual risk factors in a community sample of young adults who engage in NSSI.

In a study of NSSI and risk factors in a college sample of 133 undergraduates (89 women, 44 men), Gratz and colleagues (2002) investigated associations between NSSI, childhood trauma and abuse, and insecure attachment. The authors reported that insecure attachment, physical and sexual abuse, as well as physical and emotional neglect were significantly correlated with the frequency of NSSI as measured by the number of lifetime incidents. In addition, regression analyses indicated that maternal and paternal emotional neglect, as well as insecure attachment, significantly predicted NSSI in women; while childhood separation from caregiver predicted NSSI for men.

Subsequently, Gratz (2006) extended these earlier findings with a larger sample of women. Results revealed a relationship between early maltreatment (a composite of sexual or physical abuse and/or emotional neglect or overprotection), emotional inexpressivity, and affect intensity/reactivity to NSSI. In this study, the author concluded that childhood maltreatment and low positive affect intensity/reactivity reliably distinguished women with frequent self-injury (10 or more incidents) from those with no history of self-injury. However, maltreatment did not emerge as a significant predictor of the frequency of NSSI among the women who had engaged in NSSI, rather the frequency was better predicted by the individual risk factors (emotional inexpressivity or affect intensity/reactivity). While these two studies suggest a link between childhood maltreatment and NSSI, it is unclear which aspect of the maltreatment is contributing to the effect. Furthermore, both studies used a sample that was recruited for a study on self-injury and was limited to psychology students. Additionally, the substantial discrepancy in the reported prevalence (38% and 37%, respectively) relative to other similar samples (e.g., Favazza, DeRosear, & Conterio, 1989; Whitlock, Eckenrode, & Silverman, 2006) suggests that this sample was not fully representative of college students.

In summary, the prevalence of NSSI in community samples of high school and college students appears to be very high (10% to 20%), with greater variability in college sample estimates. In addition, although past research has suggested several risk factors linked to NSSI, the evidence of the relationship between this behavior and emotion dysregulation, attachment difficulties, and childhood sexual or physical abuse is extremely preliminary.

Research Objectives

The present study has two central goals. The first is to examine the prevalence of NSSI within a sample of college students, as well as general results regarding frequency, gender differences, method, and age of onset of NSSI. The second goal is to document risk factors associated with NSSI by comparing a group of students who engage in NSSI with a control group on variables of emotion regulation, attachment, and childhood trauma.

While previous studies have documented risk factors related to NSSI, very few studies have allocated a control group in order to perform an accurate comparison of self-injurers and the general population. While the first goal seeks to provide descriptive information concerning the occurrence of NSSI in college students, the second goal has a number of associated hypotheses. Specifically, it is hypothesized that, relative to the control group, the NSSI group will report significantly (a) higher levels of emotion dysregulation, (b) less secure attachment to their parent (or primary caregiver), and (c) more experiences of childhood trauma and abuse.

METHOD

Participants

The participants in this study were 728 students (160 male, 568 female) between the ages of 18 and 55 years (M = 20.64, SD = 3.46) at a large urban university in Montreal, Canada who agreed to complete a questionnaire on stress and coping during class time. Students were recruited from various courses in the arts (80.4%), science (8.7%), management (4.3%), music (2.2%), and education (2.2%).

Students were further classified into two groups based on their responses to a screening measure (Ross & Heath, 2002). The final sample consisted of 23 participants in the NSSI group (2 male, 21 female), ranging from 18 to 24 years in age (M = 20.22, SD = 1.76), who completed the in-class screening measure and a follow-up survey. From the remaining participants, a comparison group was selected (Control group) by generating a list of possible matches, based on gender and age, to the NSSI group. The 23 participants in the Control were randomly selected from the list of potential matches for each participant in the NSSI group (3 male, 20 female; years of age M = 20.26, SD = 1.68).

Procedure

The participants of this study were recruited through undergraduate courses at a Canadian university. Following approval from the institution's Research Ethics Board, research assistants visited large classes from various faculties across the university to request participation in a study on "how young adults deal with stress." The students were verbally informed that the purpose of the study was to explore young adults' use of various coping strategies (both adaptive and maladaptive). This involved slight deception at the onset in order to avoid any stigmatization of NSSI or a self-selected sample as noted by Gratz (2006).

Students who agreed to participate provided written consent and completed a screening measure, the HIDS (Ross & Heath, 2002), during class time. The screening was Phase I of the study, in which 728 students participated. Upon completion of the HIDS, students were given the option to provide their contact information, on a form attached at the end of each questionnaire, if they were willing to be contacted for participation in a follow-up survey (Phase II).

Participants who reported a history of NSSI behaviors and provided their contact information were e-mailed an invitation to participate in the follow-up study on stress and coping (Phase II). They were informed that participation in this follow-up would entail completion of a more in-depth survey, which would consist of three questionnaires related to attitudes towards feeling, family relationships, and childhood memories, as well as several questions regarding the coping strategies that they endorsed in the initial questionnaire that they completed (Phase I). Students were offered $20 as compensation for their time.

From an original number of 85 who had endorsed the NSSI screening item, 23 participants agreed to complete the Phase II follow-up survey. These participants were matched (by gender and age) with a control group who reported no past experience with NSSI. The final sample of 46 students (NSSI group and Control group) were e-mailed the follow-up survey; consisting of all measures described below except for the previously administered screening measure. Participants completed the follow-up survey at their convenience, and submitted the completed document through e-mail. After completion, participants were e-mailed debriefing information and a leaflet containing resources and information for individuals who engage in NSSI.

Measures

How I Deal with Stress Questionnaire (Ross & Heath, 2002). This questionnaire, designed and reported by Ross and Heath (2002), was employed as a screening measure. This measure presented 25 coping strategies and participants were asked to indicate how often they employed this strategy on a 4-point Likert scale (never, once, a few times, frequently). Nonsuicidal self-injury (NSSI) was embedded within this questionnaire as one of the possible coping strategies. Follow-up sections were included for those individuals who indicated that, when feeling stress, they "talk to someone," "do risky things," or "hurt oneself on purpose." These sections asked students to provide more details about their use of this strategy. Specifically, for those who indicated that they had hurt themselves on purpose, they were asked to specify the method (e.g., cutting, burning), frequency of the behavior, and the intention of the behavior (whether they had ever self-injured with suicidal intent).

No validity and reliability information is available for this measure as it was designed by Ross and Heath (2002) for use in a study examining NSSI in a community sample of adolescents. However, research in this field often employs questionnaires designed by the investigators to identify and assess the extent of self-injurious behavior (e.g., Dulit, Fryer, Leon, Brodsky, & Frances, 1994; Favazza & Conterio, 1989). In addition, the questionnaire serves only to detect the presence or absence of NSSI within this sample. This measure was selected for screening, as opposed to other measures that have been used to tap acts of self-harm (e.g., Deliberate Self-Harm Inventory; Gratz, 2001), in order to avoid the selection bias which may have occurred were it announced to all screened individuals that this was a study specifically examining NSSI.

Deliberate Self-Harm Inventory (Gratz, 2001). The Deliberate Self-Harm Inventory (DSHI) is a 17-item, behaviorally based, self-report questionnaire developed by Gratz (2001) to assess self-injurious behaviors. The DSHI is based on the conceptual definition of NSSI as the deliberate, direct destruction or alteration of body tissue without conscious suicidal intent, but resulting in injury severe enough for tissue damage to occur. The DSHI has been shown to have adequate internal consistency ([alpha] = .82), as well as adequate test-retest reliability over a period ranging from 2 to 4 weeks, with a mean of 3.3 weeks (r = .68,p < .001). The DSHI has also shown good construct, discriminant, and convergent validity (Gratz, 2001). This measure was completed by the NSSI group only, in Phase II of the study.

Difficulties in Emotion Regulation Scale (Gratz & Roemer, 2004). The Difficulties in Emotion Regulation Scale (DERS) is a 36-item questionnaire which assesses six dimensions of emotion regulation: personal acceptance or denial of emotional responses (Non-Acceptance), ability to engage in daily life or accomplish tasks when overwhelmed with emotion (Goals), ability to control emotions and reactions to these emotions (Impulse), awareness and acknowledgement of emotions (Awareness), access and employment of tools to regulate emotions (Strategies), and ability to understand and interpret personal emotions (Clarity). Participants were asked to indicate, on a five-point Likert scale ranging from "almost never" to "almost always," how often each statement was true for them. Gratz and Roemer (2004) found this measure to have high internal consistency ([alpha] = .93), good test-retest reliability over a period ranging from 4 to 8 weeks, and adequate construct validity (r = .60) when correlated with the Acceptance and Action Questionnaire (Hayes, Wilson, Gifford, Follette, & Strosahl, 1996). In the current study, internal consistency as measured by Cronbach's alpha was high ([alpha] = .96).

Reciprocal Attachment Questionnaire (West & Sheldon-Keller, 1992; 1994). To assess attachment, the Reciprocal Attachment Questionnaire (RAQ), developed and published by West and Sheldon-Keller (1992; 1994) was used. The 15-item version of the RAQ published in 1994 was employed, assessing five dimensions of attachment in adults: availability and responsiveness, feared loss, proximity seeking, separation protest, and use of attachment figure. Psychometric properties of the complete 35-item version of this measure include adequate construct validity as tested with factor analysis, and good reliability, with scales' internal consistency reliabilities ranging between .79 and .88 (West & Sheldon-Keller, 1992). The instrument was tested with both psychiatric and community samples. The computed Cronbach's alpha was .76 in present study, indicating reasonable internal consistency reliability.

Childhood Trauma Questionnaire (Bernstein & Fink, 1998). This questionnaire developed by Bernstein and Fink (1998) is a retrospective measure assessing traumatic childhood experiences along five dimensions: emotional, physical, sexual abuse, emotional, and physical neglect. This 25-item version was derived from a longer version containing 70 items which demonstrated high levels of test-retest reliability (interclass correlation = .88), and internal consistency with Cronbach's alpha between .79 and .94 (Bernstein et al., 1994). In addition, it was tested with widely different samples: adult substance abusing patients from New York City, adolescent psychiatric inpatients, substance abusing individuals from southwest Texas and persons from the community. The results indicated good criterion-validity, as the instrument's constructs significantly predicted the observational scores by the therapists, with regression standardized coefficients ranging from .24 to .27 (Bernstein et al., 2003). For the current study, internal consistency as measured by Cronbach's alpha was reasonable ([alpha] = .82).

RESULTS

Prevalence of Nonsuicidal Self-Injury

The occurrence of self-injurious behaviors amongst this sample of college students was calculated using the screening data from the HIDS. From the sample of 728, 85 participants indicated that they "hurt themselves on purpose" without suicidal intent in the screening questionnaire, representing an 11.68% prevalence rate of NSSI. From the follow-up questions on the HIDS, the frequency of NSSI behaviors was examined and the distribution was as follows: 23.6% once, 40.3% two to four times, 8.3% five to ten times, 16.7% 11 to 50 times, 2.8% 51 to 100 times and 4.2% over 100 times.

Of the 85 students, 15 were male (9.4% of all male participants) and 70 were female (12.3% of all female participants). A Pearson Chi-Square test was used to examine a possible difference in prevalence between men and women. There was no significant difference found, [chi square] (2) = 4.581, p = n.s. Significantly fewer male students were willing to complete follow up interviews (a Pearson Chi-Square test was interpreted with a more lenient alpha level, [chi square] (1) = 2.855, p = .091). Only 33.3% of the male students who admitted to engaging in NSSI in the screening (Phase I) agreed to be complete follow-up survey (Phase II), in contrast to the 58.6% of female self-injurers who agreed to complete the follow-up.

Characteristics of Reported Nonsuicidal Self-Injury

The remaining analyses represent only the participants who completed Phase II of the study. To further investigate the general responses provided by the NSSI group (n = 23), frequencies and means were calculated for several responses on the Deliberate Self-Harm Inventory (DSHI). Participants reported engaging in various forms of NSSI including cutting (65.2%), severe scratching (56.5%), punching self (26.1%), burning (21.7%), and banging head (8.7%). In terms of age of onset, the majority of participants (43.5%) reported their first incident of NSSI between the age of 14 and 16 years, 18.3% between 11 to 13 years, 21.7% reported their first incident between 17 to 19 years, and 17.3% over the age of 20 years (M = 16.09; SD = 3.16).

Frequency was measured by participants' reports of the number of incidents with NSSI over a lifetime. One incident was reported by 8.7%, 26.1% reported two to four incidents, 47.8% percent reported five to ten incidents, 8.7% reported 11 to 50 incidents, 4.3% reported 51 to 100 incidents, and 4.3% reported over 100 incidents. Intensity was measured by reports of the number of times the NSSI occurred during periods when they were self-injuring. Responses were grouped by one to two times ever (27.3%), once per month (40.9%), one to four times per month (18.2%), and two to seven times per week (13.6%).

Investigation of Risk Factors

To investigate possible between-group differences in reports of emotion regulation difficulties, attachment, and childhood abuse, Multivariate Analysis of Variance (MANOVA) was employed. Specifically, three separate MANOVAs were conducted to test for significant differences between the NSSI group and the Control group on emotion regulation (DERS subscales), attachment (RAQ subscales), and childhood abuse (CTQ subscales). Total scale scores were not included in these analyses as they are composites of the subscales. The descriptive statistics for each of these measures across the NSSI and Control groups is provided in Table 1.

In examining the results for the DERS, a significant difference was found, Wilk's [LAMBDA] = .59, F (6, 39) = 4.46, p = 0.002, multivariate [[eta].sup.2] = .41, which is a large effect size according to Cohen (1988). Observed power was also high (.64). Examination of the coefficients for the linear combination distinguishing groups indicated that the Strategies (-6.81), Impulse (-5.96), and Goals (-4.44) subscales contributed most to the difference in emotion regulation. Further examination of the univariate effects indicated that five of the six subscales were significantly different for the NSSI group and Control group. Between-subject effects for the DERS are presented in Table 2.

Two additional MANOVAs were conducted to test for between-group differences on reports of attachment and childhood abuse. No significant differences were revealed for the analyses of the RAQ, Wilk's [LAMBDA] = .90, F (5, 40) = .87, p = n.s., multivariate [[eta].sup.2] =. 10, nor for the CTQ, Wilk's [LAMBDA] = .89, F (5, 40) = .98, p = n.s., multivariate [[eta].sup.2] =. 11. However, it is important to note the observed power for the RAQ and CTQ analyses were low, .28 and .31, respectively.

Follow-up analyses were conducted to examine whether risk factors scores for the NSSI group would be related to the number of lifetime incidents of self-injury. Bivariate Pearson correlations were run to test the relationship between the total scale score of the DERS, RAQ, and CTQ and frequency, as determined by participants' reports of how many times he or she had engaged in NSSI over his or her lifetime (once, 2-4, 5-10, 11-50, 51-100, 100+ times). Results revealed no significant correlations between lifetime frequency and total scores on the DERS (r = .25, p = .25), RAQ (r = -.33, p = .12), or CTQ (r = -.22, p = .32).

DISCUSSION

The present study had two objectives; first, to examine the characteristics of NSSI within a sample of college students, including prevalence, frequency, gender differences, method, and age of onset. The second objective was to compare a group of students who engaged in NSSI with a comparison non-NSSI group on identified risk factors (emotion regulation, attachment, and childhood trauma).

Results concerning the characteristics of NSSI in this sample were informative in light of previous literature, in that many of the findings were consistent with past research. The reported age of onset for NSSI behaviors was primarily in early adolescence. Not surprisingly, the most common reported method of NSSI was cutting of arms or other areas of the body, followed by scratching of the skin, self-punching, and burning. Interestingly, the prevalence of NSSI as reported in the initial screening measure (Phase I), did not differ by gender. Specifically, 12% of female students and 9.4% of male students admitted to engaging in NSSI at least once. Significantly fewer men were willing to complete the Phase II follow-up survey and, as a result, gender comparisons were not possible between the NSSI group and Control group. However, the reported prevalence rates further support the notion that gender differences are increasingly not being found in the occurrence of NSSI within community samples.

In contrast to the 38% prevalence rates reported by Gratz et al. (2002) and the 37% rated reported by Gratz (2006), the current investigation revealed a prevalence rate of 11.68% for NSSI. One obvious explanation for this discrepancy is the difference in sample selection. The former studies by Gratz and colleagues (2002; 2006) recruited participants for a study on self-injurious behavior. The participants in the current sample were told that this was a study was about "how young adults deal with stress" and coping behaviors. Furthermore, students were drawn from various faculties and programs across the university, and not limited to psychology students. Another explanation for the differing prevalence rates may be measure sensitivity. In the Gratz studies, the questionnaire completed by the participants provided a comprehensive checklist of all possible NSSI behaviors and they were asked to indicate if they had ever engaged in any of these listed behaviors. In the current study, the screening measure asked participants if they had ever "hurt themselves on purpose" and thus, it is possible that some individuals may not have recalled behaviors that were tapped in the checklist, such as biting, rubbing skin excessively, or interfering with wound healing. Future studies are needed that employ both types of measurement of NSSI behavior.

Consistent with Whitlock, Eckenrode and Silverman's (2006) results concerning age of onset, the majority of participants in the present study first engaged in NSSI during adolescence, however, a substantial number reported engaging in this behavior for the first time after the age of 17 years (39%). Whitlock and colleagues had 38.6% of participants reporting an age onset between 17-24 years. This finding clearly indicates that a significant portion of those who will engage in NSSI will begin to do so during their college years. Further research is needed to investigate the factors that contribute to the onset of this behavior and particularly, why individuals initiate this behavior as young adults versus as adolescents. Furthermore, this finding suggests an important possibility for college and community mental health services, in that they have the opportunity to develop preventative programs, as well as treatment programs.

Another informative result was found in relation to the frequency of NSSI behavior reported by participants. In the sample of 85 students who reported NSSI (11.68% of the total sample), the majority had engaged in this behavior less than ten times, with 23.6% indicating that they self-injured only once. Thus, the present sample tapped a group of individuals who are largely not repetitive self-injurers. Nevertheless, it is of interest and import to understand that there is a significant number of young adults in this high-achieving community sample (e.g., enrolled in a competitive international university) that are engaging in NSSI minimally or as a seemingly transient behavior. Understanding that this behavior is occurring is essential to working towards distinguishing between individuals who self-injure repetitively and those who engage in NSSI during adolescence or young adulthood, similar to many other high-risk behaviors. Thus, the present results suggest that more severe, repetitive forms of NSSI remain infrequent amongst college students, although less severe forms of NSSI appear to be relatively common.

The exploration of risk factors revealed a significant difference between the NSSI and Control groups in ratings of emotion regulation difficulties. This is particularly interesting that emotion regulation was found to play such a central role, even among this group that does not engage in repetitive forms of NSSI. Furthermore, it was found that difficulties with emotion regulation exist among the NSSI group regardless of the frequency of this behavior (as measured by number of lifetime incidents). The emotion regulation subscales that emerged as most problematic for the NSSI group were Strategies, Impulse, and Goals. This finding indicates that these individuals do not have a repertoire of strategies to employ when they are dealing with stress or difficulty. Impulse and Goals are related to an individual's ability to control the reactions to their emotions, as well as continue to function when experiencing strong emotions.

The lack of significant group differences on measures of attachment (RAQ) or childhood trauma (CTQ) was contrary to the hypotheses. It may be that these risk factors were not found in this sample because the participants engaged in a mild form of NSSI. Attachment difficulties and childhood trauma may be found mainly in clinical samples or amongst young adults with a higher lifetime frequency of this behavior. However, these findings require replication with a larger sample in light of the moderate effect sizes and low levels of observed power in the present investigation.

Despite this caution in interpreting the present results, it is relevant for practitioners to note the substantial difference between self-injurers and non-self-injurers in emotion regulation; even in the absence of chaotic family histories (e.g., attachment difficulties, childhood trauma or abuse). Past literature and clinical reports have frequently documented childhood trauma and abuse, characteristic of severe self-injurers, as being causally linked to the observed emotion regulation difficulties (e.g., Linehan, 1993). It has been believed that dysfunctional or chaotic family histories have resulted in poor emotion regulation which the individual manages with maladaptive behaviors, such as NSSI. The fact that emotion regulation difficulties were present in this sample, in the absence of these other risk factors, suggests that emotion dysregulation is a central difficulty for a college sample of self-injurers.

Implications for Mental Health Counselors

It is critical for mental health counselors working with young adults who engage in NSSI to be aware that low frequency occurrences of this behavior are quite common, despite relatively low rates of severe or repetitive NSSI. This information has important implications for referral to support or treatment services within college and community settings. Currently, school- and community-based mental health professionals are often unsure of how to proceed with incidents of NSSI and psychiatric services observe an increase in referrals of individuals who engage in this behavior. Paris (2005) notes that much of the literature on NSSI in the community has failed to distinguish between repetitive and occasional NSSI; and that an overreaction to a single or occasional occurrence of NSSI can result in unnecessary hospitalization. In addition, the current thinking in the field regarding the perceived increase in this behavior has failed to evaluate whether the rise in this behavior is limited to the single or more occasional occurrence of NSSI rather than the chronic repetitive NSSI.

However, this is not to suggest that NSSI at the frequency reported in the current sample is of no concern. First, there is some accumulating evidence that NSSI may have addictive properties for some individuals (e.g., Nixon, Cloutier, & Aggarwal, 2002) and the more individuals who "try" this behavior, the more who may be at risk for repetitive NSSI. Second, as the current study demonstrates, even those who are not repetitive self-injurers still have significantly greater problems with emotion regulation. Specifically, they have problems controlling their reactions to strong emotions or functioning when experiencing these emotions. Perhaps the most problematic was the perceived lack of strategies to regulate their emotions. Muehlenkamp's (2006) critical review of effective interventions for NSSI highlights the need to focus on developing strategies to manage difficulties with emotion regulation.

Finally, the current study assessed the frequency of NSSI incidents, but did not assess severity of the incidents. This may be an important distinction and even occasional or infrequent NSSI that is severe (e.g., deep cutting requiring medical attention, severe burns, or multiple injuries in each incident) must be dealt with as severe NSSI.

The finding that NSSI and attendant emotion regulation difficulties clearly occur in the absence of attachment difficulties and childhood trauma should serve as a "red flag" to practitioners not to assume the presence of early abuse when confronted with a client engaging in NSSI. The link between early abuse and NSSI that has been a robust finding in studies with clinical samples is far more tenuous in college samples. Adler and Adler (2007) completed an ethnographic, interpretive inquiry into the experience of individuals who engage in NSSI. Following eighty in-depth interviews with individuals who engage in NSSI, the authors identified four apparent groups: (a) the traditional inpatient populations, (b) the poor, weak or powerless, (c) older long term self-injurers, and (d) the "mildly disturbed, alienated, or typically angst-ridden teenagers and young adults" (p. 559). In examining the authors' analyses of the interviews, the last group appears to be consistent with many of the participants in the present study. Interestingly, participants from all groups talked about how NSSI helps them to regulate or manage their emotions, making their emotional pain bearable and visible. Practitioners working with individuals who engage in NSSI need to focus on helping the client to find more adaptive ways to tolerate intense emotions and regulate their emotions.

Limitations

While the results of the present study represent a valuable addition to the current literature base, there are nonetheless some limitations. First, the information was gathered using self-report measures. Although this is the method used by most researchers investigating self-injurious behavior (e.g., Favazza & Conterio, 1988; Favazza & Conterio, 1989; Gratz et al., 2002; Gratz, 2001; Whitlock, Eckenrode, & Silverman, 2006), it could also present some potential problems such as social desirability and thus, underreporting of the behavior. In the present study, efforts have been made to increase participants' comfort in disclosing their experiences by not describing the focus of the study as NSSI and making all reports fully anonymous through large-scale survey (Phase I) and electronic surveys (Phase II). However, it is possible that some individuals may have distorted the information they reported. A second limitation is related to the generalizability of the results; because the participants in the NSSI group (Phase II) were predominantly female, the present findings cannot be generalized to male students. In addition, the use of an adaptation of the RAQ may have weakened the validity of the measure. Future work needs to be done using a retrospective measure of attachment with better psychometric properties. Finally, the relatively small number of participants in Phase II limits the statistical power to detect differences and thus, while the group differences in emotion regulation are robust, the lack of a significant difference on the attachment and childhood trauma measures needs to be examined in future studies in light of the moderate effect sizes and limited power.

CONCLUSION

In summary, the current study found NSSI to be quite prevalent amongst a sample of college students. In anonymous screening reports, there was found to be no gender difference in the prevalence of NSSI behaviors. However, significantly fewer men were willing to complete the follow-up study. This finding may be mirrored by reluctance on the part of young men to seek help for this behavior, which could be falsely interpreted as a lower prevalence of NSSI amongst men. In addition, lifetime frequency of NSSI was not related to emotion regulation difficulties, suggesting that even those who engage in non-repetitive NSSI suffer from difficulties in managing their emotions. Interestingly, no significant difference was found between those who self-injured and those who did not in reports of attachment difficulties, childhood trauma or abuse. Therefore, it is apparent that emotion dysregulation and NSSI can occur in the absence of dysfunctional or chaotic family histories.

In conclusion, the present study provides some interesting and provocative information for practitioners in the area. The need for distinguishing between "mild" NSSI and more severe forms when documenting prevalence is accentuated. The lack of support for the hypotheses that those who engaged in NSSI would differ from non-NSSI peers on childhood trauma was unexpected and if replicated, indicates a future direction in differentiating mild and severe NSSI based on risk factors. Finally, the difficulties in emotion regulation reported by the participants who engaged in even mild NSSI are intriguing. Emotion regulation appears to be central to the occurrence of this behavior in all its forms. Clearly, differentiating between potential subtypes of NSSI both clinically and empirically is essential.

REFERENCES

Adler, P. A., & Adler, P. (2007). The demedicalization of self-injury: From psychopathology to sociological deviance. Journal of Contemporary Ethnography, 36, 537-550.

Ayton, A., Rasool, H., & Cottrell, D. (2003). Deliberate self-harm in children and adolescents: Association with social deprivation. European Child & Adolescent Psychiatry, 12, 303-307.

Bernstein, D. P., Stein, J. A., Newcomb, M. D., Walker, E., Pogge, D., Ahluvalia, T., et al. (2003). Development and validation of a brief screening version of the Childhood Trauma Questionnaire. Child Abuse and Neglect, 27, 169-190.

Bernstein, D., & Fink, L. (1998). Childhood Trauma Questionnaire. A Retrospective Self-Report. [Manual]. San Antonio: Harcourt Brace & Company.

Bernstein, D. P., Fink, L., Handelsman, L., Foote, J., Lovejoy, M., Wnzel, K., Sapareto, E., & Ruggiero, J. (1994). Initial reliability and validity of a new retrospective measure of child abuse and neglect. The American Journal of Psychiatry, 151, 1132-1136.

Boudewyn, A. C., & Liem, J. H. (1995). Childhood sexual abuse as a precursor to depression and self-destructive behavior in adulthood. Journal of Traumatic Stress, 8, 445-459.

Briere, J., & Gil, E. (1998). Self-mutilation in clinical and general population samples: Prevalence, correlates, and functions. American Journal of Orthopsychiatry, 64, 609-520.

Carroll, J., Schaffer, C., Spensley, J., & Abramowitz, S. I. (1980). Family experiences of self-mutilating patients. American Journal of Psychiatry, 137, 852-853.

Cohen, J. (1988). Statistical power analysis for the behavioral sciences (2nd ed.). Hillsdale, N J: Erlbaum.

Conterio, K., & Lader, W. (1998). Bodily harm: The breakthrough healing program for self-injurers. New York: Hyperion.

D'Onofrio, A. A. (2007). Adolescent sell-injury: A comprehensive guide for counselors and health care professionals. New York: Springer Publishing.

Dulit, R. A., Fryer, M. R., Leon, A. C., Brodsky, B. S., & Frances, A. J. (1994). Clinical correlates of self-mutilation in borderline personality disorder. The American Journal of Psychiatry, 151, 1305-1312.

Favazza, A. R. (1998). The coming age of self-mutilation. Journal of Nervous & Mental Disease, 186, 259-268.

Favazza, A. R., & Conterio, K. (1989). Female habitual self-mutilators. Acta Psychiatrica Scandinavica, 79, 283-289.

Favazza, A., & Conterio, K. (1988). The plight of chronic self-mutilators. Community Mental Health Journal 24, 22-30.

Favazza, A. R., DeRosear, L., & Conterio, K. (1989). Self-mutilation and eating disorders. Suicide and Life-Threatening Behavior, 19, 352-361.

Gratz, K. L. (2006). Risk factors for deliberate self-harm among female college students: The role and interaction of childhood maltreatment, emotional inexpressivity, and affect intensity/reactivity. American Journal of Orthopsychiatry, 76, 238-250.

Gratz, K. (2001). Measurement of deliberate self-harm: preliminary data on the deliberate self harm inventory. Journal of Psychopathology and Behavioral Assessment, 23, 253-263.

Gratz, K., Conrad, S. D., & Roemer, L. (2002). Risk factors for deliberate self-harm among college students. American Journal of Orthopsychiatry, 72, 128-140.

Gratz, K., & Roemer, L. (2004). Multidimensional assessment of emotional regulation and dysregulation: development, factor structure, and initial validation of the Difficulties in Emotion Regulation Scale. Journal of Psychopathology and Behavioral Assessment, 26, 41-54.

Haines, J., Williams, C. L., Brain, K. L., & Wilson, G. V. (1995). The psychophysiology of self-mutilation. Journal of Abnormal Psychology, 104, 471-489.

Hawton, K., Fagg, J., Simkin, S., Bale, E., & Bond, A. (2000). Deliberate self-harm in adolescents in Oxford, 1985-1995. Journal of Adolescence, 23, 47-55.

Hawton, K., Hall, S., Simkin, S., Bale, L., Bond, A., Codd, S., et al. (2003). Deliberate self-harm in adolescents: A study of characteristics and trends in Oxford, 1990-2000. Journal of Child Psychology and Psychiatry, 44, 1191-1198.

Hawton, K., & James, A. (2005). Suicide and deliberate selfharm in young people. BMJ,, 330, 891-894.

Hawton, K., Rodham, K., Evans, E., & Weatherall, R. (2002). Deliberate self-harm in adolescents: Self-report survey in schools in England. British Medical Journal, 325, 1207-1211.

Hayes, S. C., Wilson, K. G., Gifford, E. V., Follette, V. M., & Strosahl, K. (1996). Experiential avoidance and behavioral disorders: A functional dimensional approach to diagnosis and treatment. Journal of Consulting and Clinical Psychology, 64, 1152-1168.

Heath, N. L., Toste, J. R., & Beet-tam, E. (2007, November). School counsellors 'experiences with self-injury in the schools. Paper presented at the Canadian Academy for Child and Adolescent Psychiatry annual conference, Montreal, Quebec.

Heath, N. L., Toste, J. R., & Beettam, E. (2006). "I am not well-equipped": High school teachers' perceptions of self-injury. Canadian Journal of School Psychology, 21, 73-92.

International Society for the Study of Self-injury (2007). Definitional issues surrounding our understanding of sell-injury. Conference proceedings from the annual meeting.

Izutsu, T., Shimotsu, S., Matsumoto, T. Okada, T., Kikuchi, A., Kojimoto, M., et al. (2006). Deliberate self-harm and childhood hyperactivity in junior high school students. European Child and Adolescent Psychiatry, 14, 1-5.

Jacobsen, C. M., Muehlenkamp, J. J., & Miller A. L. (2006). Psychiatric impairment among adolescents engaging in different types of deliberate self-harm. Manuscript submitted for publication.

Klonsky, E. D. (2007). The functions of deliberate NSSI: A review of the evidence. Clinical Psychology Review, 27, 226-239.

Laye-Gindhu, A., & Shonert-Reichl, K. (2005). Nonsuicidal self-harm among community adolescents: Understanding the "whats" and "whys" of self-harm. Journal of Youth and Adolescence, 34, 445-457.

Linehan, M. M. (1993). Cognitive-behavioral treatment of borderline personality disorder. New York: The Guilford Press.

Lloyd-Richardson, E. E., Perrine, N., Dierker, L., & Kelley, M. L. (2007, in press). Characteristics and functions of nonsuicidal self-injury in a community sample of adolescents. Psychological Medicine.

Muehlenkamp, J. J. (2006). Empirically supported treatments and general therapy guidelines for nonsuicidal self-injury. Journal of Mental Health Counseling, 28, 166-185.

Muehlenkamp, J. J., & Gutierrez, E M. (2004). An investigation of differences between self-injurious behavior and suicide attempts in a sample of adolescents. Suicide and Life-Threatening Behavior, 34, 12-23.

Muehlenkamp, J. J., & Gutierrez, P. M. (2007). Risk for suicide attempts among adolescents who engage in nonsuicide self-injury. Archives of Suicide Research, 11, 69-82

Nixon, M. K., Cloutier, P. F., & Aggarwal, S. (2002). Affect regulation and addictive aspects of repetitive self-injury in hospitalized adolescents. Journal of the American Academy of Child & Adolescent Psychiatry, 41, 1333-1341.

Nixon, M. K., Cloutier, P., & Jansson, S. M. (2007). Prevalence, mental health correlates and help seeking for nonsuicidal self-harm in a sample of Canadian youth. Manuscript submitted for publication.

Paris, J. (2005). Understanding self-mutilation in borderline personality disorder. The Harvard Review Psychiatry, 13, 179-185.

Patton, G. C., Harris, R., Carlin, J. B., Hibbert, M. E., Coffey, C., Schwartz, M., et al. (1997). Adolescent suicidal behaviors: A population-based study of risk. Psychological Medicine, 27, 715-724.

Rodham, K., Hawton, K., & Evans, E. (2004). Reasons for deliberate self-harm: Comparison of self-poisoners and self-cutters in a community sample of adolescents. Journal of the American Academy of Child and Adolescent Psychiatry, 43, 80-87.

Ross, S., & Heath, N. L. (2002). A study of the frequency of self-mutilation in a community sample of adolescents. Journal of Youth and Adolescence, 31, 67-77.

Shaw, S. N. (2002). Shifting conversations on girls' and women's self-injury: An analysis of the clinical literature in historical context. Feminism & Psychology, 12, 191-219.

Shields, A., & Cicchetti, D. (1998). Reactive aggression among maltreated children: The contributions of attention and emotion dysregulation. Journal of Clinical Child Psychology, 27, 381-395.

Silverman, M., Berman, A., Sanddal, N., O'Carraoll, P.W., & Joiner, Thomas, E., M., et al. (2007). Rebuilding the tower of Babel: A revised nomenclature for the study of suicide and suicidal Behaviors--Part 2: Suicide-related ideations, communications, and behaviors. Suicide and Life-Threatening Behavior, 37, 264-277.

Suyemoto, K. L. (1998). The Functions of Self-Mutilation. Clinical Psychology Review, 18, 531-554.

Tulloch, A. L., Blizzard, L., & Pinkus, Z. (1997). Adolescent-parent communication in self- harm. Journal of Adolescent Health, 21, 267-275.

van der Kolk, B. A. (1996). The complexity of adaptation to trauma: Self-regulation, stimulus discrimination, and characterological development. In B. A. van der Kolk, A. C. McFarlane, & L. Weisaeth (Eds.), Traumatic stress: The effects of overwhelming experience on mind, body, and society (pp. 182-213). New York: Guilford Press.

van der Kolk, B. A., & Fisler, R. E. (1994). Childhood abuse and neglect and loss of self-regulation. Bulletin of the Menninger Clinic, 58, 145-168. Walsh, B. W. (2006). Treating self-injury: A practical guide. New York: Guilford Press.

West, M., & Sheldon-Keller, A. E. (1994). Patterns of relating. An adult attachment perspective. New York: The Guilford Press.

West, M., & Sheldon-Keller, A. (1992). The assessment of dimensions relevant to adult reciprocal attachment. Canadian Journal of Psychiatry, 37, 600-605.

Whitlock, J., Purington, A., Eells, G., & Cummings, N. (2006). Self-injurious behavior in college populations: Perceptions and experiences of college mental health providers. Manuscript submitted for publication.

Whitlock, J., Eckenrode, J., & Silverman, D. (2006). Self-injurious behaviors in a college population. Pediatrics, 114, 1939-1948.

Zila, L. M., & Kiselica, M. S. (2001). Understanding and counseling self-mutilation in female adolescents and young adults. Journal of Counseling & Development, 79, 46-52.

Zoroglu, S. S., Tuzun, U., Sar, V., Tutkun, H., Savas, H. A., Ozturk, M., et al. (2003). Suicide attempt and self-mutilation among Turkish high school students in relation with abuse, neglect and dissociation. Psychiatry & Clinical Neurosciences, 57, 119-126.

Nancy Heath, Jessica Toste, Tatiana Nedecheva and Alison Charlebois are affiliated with the Department of Educational and Counselling Psychology at McGill University in Montreal, Canada. Correspondence regarding this article may be directed to Nancy Heath. E-mail: [e-mail].

Table 1 Means and Standard Deviations on the DERS, RAQ, and CTQ for
the NSSI and Control Groups

                               Group designation

                         NSSI group     Control group
                          (n = 23)       (n = 23)

Variables               M       SD       M       SD

Difficulties in Emotional Regulation Scale (DERS)

  Total DERS score    101.31   26.14   78.43   21.29
  Non-Acceptance       17.04    6.48   13.09    5.95
  Clarity              12.68    3.65   10.57    3.12
  Goals                19.22    4.37   14.78    4.85
  Impulse              15.91    5.81    9.96    3.07
  Awareness            13.17    5.02   13.56    4.58
  Strategies           23.29    6.23   16.48    6.79

Reciprocal Attachment Questionnaire (RAQ)

  Total RAQ score      36.28   10.19   32.48    7.91
  Proximity Seeking     9.13    3.08    8.57    2.92
  Separation Protest    7.20    2.35    6.26    2.49
  Feared Loss           6.57    3.37    5.09    3.07
  Availability          6.91    3.41    5.91    2.35
  Use                   6.48    3.41    6.65    3.47

Childhood Trauma Questionnaire (CTQ)

  Total CTQ score      34.22    7.79   32.43    5.96
  Emotional Abuse       8.15    2.63    7.48    2.43
  Physical Abuse        5.61    1.18    5.26    0.45
  Sexual Abuse          5.65    1.56    5.43    1.24
  Emotional Neglect     8.37    3.36    8.43    2.92
  Physical Neglect      6.43    1.55    5.83    1.11

Table 2. MANOVA Between-Subject Effects on DERS Subscales

Dependent variable   F(1, 46)   H      p

Non-Acceptance         4.65     .10   .037
Clarity                4.48     .09   .040
Goals                 10.61     .19   .002
Impulse               18.92     .30   .000
Awareness               .08     .00   .784
Strategies            12.55     .22   .001