84

Suicide and Life-Threatening Behavior 45 (1) February 2015 ©2014 The American Association of Suicidology DOI: 10.1111/sltb.12117

Factors Associated with Current Versus Lifetime Self-Injury Among High School and College Students LINDSAY A. TALIAFERRO, PHD, MPH,

AND

JENNIFER J. MUEHLENKAMP, PHD

We sought to identify factors associated with current versus lifetime nonsuicidal self-injury (NSSI) and factors that show consonant and distinct relationships with current NSSI for adolescents and young adults. Data came from a population-based survey of high school students (n = 9,985) and a national survey of college students (n = 7,801). Among both samples, factors associated with current NSSI included male gender, younger age, greater depressive symptoms, more hopelessness, and being the victim of a verbal or physical assault. For high school students, greater anxiety, and for college students, identifying as non-White, negative perceptions of one’s weight, a same-sex sexual experience, and involvement in dating violence also distinguished the groups. Findings suggest that clinical and research assessments of lifetime NSSI might not extend to current behavior, and some differences exist in the factors associated with current behavior between adolescents and young adults. Clinical practice and prevention programming efforts should target certain intrapersonal and interpersonal factors associated with current NSSI among younger students during stressful transition periods in their lives, such as entering high school or college, when they might consider initiating or continuing this behavior. Nonsuicidal self-injury (NSSI) among young people, defined as purposely inflicting injury that results in immediate tissue damage, done without suicidal intent, and not socially sanctioned (Klonsky, Muehlenkamp, Lewis, & Walsh, 2011), represents an important public health issue. Lifetime prevalence estimates suggest 18% of high LINDSAY A. TALIAFERRO, Department of Health Sciences and Human Performance, College of Natural and Health Sciences, University of Tampa, Tampa, FL, USA; JENNIFER J. MUEHLENKAMP, Psychology Department, University of Wisconsin-Eau Claire, Eau Claire, WI, USA. Address correspondence to Lindsay Taliaferro, PhD, Department of Health Sciences and Human Performance, College of Natural and Health Sciences, University of Tampa, 401 W. Kennedy Blvd., Tampa, FL 33606; E-mail: [email protected]

school-aged youth (Muehlenkamp, Claes, Havertape, & Plener, 2012) and 12% to 38% of college students (Heath, Toste, Nedecheva, & Charlebois, 2008; Whitlock, Eckenrode, & Silverman, 2006; Whitlock et al., 2011) have engaged in NSSI. Although NSSI remains distinct from a suicide attempt, self-injury increases risk of suicidal behavior (Joiner, Ribeiro, & Silva, 2012; Whitlock et al., 2013). Despite the prevalence and risk for suicide, few population-based surveys include measures of NSSI, so researchers and clinicians possess limited knowledge about epidemiologic correlates of the behavior among young people. Within the research that exists, investigators found relationships between NSSI and sexual orientation, history of abuse, psychological distress, disordered eating/dieting behavior,

TALIAFERRO

AND

MUEHLENKAMP

substance use, interpersonal violence, and trying inhalants (Alfonso & Dedrick, 2010; Hilt, Nock, Lloyd-Richardson, & Prinstein, 2008; Taliaferro, Muehlenkamp, Borowsky, McMorris, & Kugler, 2012; Whitlock & Knox, 2007; Whitlock et al., 2011). However, investigators have not compared effects of different factors on NSSI for diverse age groups during adolescence. Most research in this area comes from relatively small community samples, and investigators have not compared risk correlates associated with NSSI among adolescents and young adults in the same study. Thus, researchers and practitioners lack complete understanding about the most important factors that predict NSSI, and whether certain factors demonstrate similar significance for adolescents as well as young adults. Given that much of the current understanding of NSSI comes from smaller samples, comparing high school and college students from national or population-based samples will help clarify whether clinicians can assess the same risk factors among adolescents and young adults; or, whether they should assess different factors when working with individuals comprising these distinct developmental groups. Furthermore, most research addressing NSSI among young people has incorporated measures of the behavior that assess lifetime incidents rather than current self-injury (Klonsky et al., 2011; Taylor, Peterson, & Fischer, 2012), and few researchers have compared diverse correlates of current versus lifetime NSSI. Within the extant literature, investigators focused on psychological characteristics and found that current self-injurers were more likely than those with a history of the behavior to report affect regulation motives for NSSI and current negative affect (Taylor et al., 2012); less acceptance of personal emotional responses and less impulse control (Anderson & Crowther, 2012); greater frequency of self-injury, lower cognitive reappraisal, and higher emotional suppression (Andrews, Martin, Hasking, & Page, 2013);

85 and higher levels of hostility, guilt, and sadness (Brown, Williams, & Collins, 2007). In addition, Rotolone and Martin (2012) found that young people who currently engaged in self-injury were more likely than those with a history of the behavior to report lower levels of family support, self-esteem, resilience, and satisfaction with life, with resilience demonstrating the strongest effect. Thus, correlates of NSSI may differ depending on the time frame used to assess NSSI, suggesting we should differentiate between youth who report current versus historical self-injury on certain measures to glean an accurate understanding of the behavior. Identifying correlates related to current NSSI and examining whether or not they significantly differ from lifetime acts is essential to providing clinicians with accurate information that informs the assessment and treatment of this behavior, as well as to informing theoretical models of risk. Determining whether factors other than those previously examined show significant differences between lifetime and current NSSI also remains important for advancing research on this behavior because differences could have implications for the generalizability of results when researchers use samples that include recent or lifetime behavior. Thus, examining potential differences in risk factors among those reporting current versus lifetime NSSI will provide further information about which correlates of NSSI researchers and practitioners would want to consider when assessing and addressing the behavior among adolescents and young adults. The current investigation sought to address gaps in the literature on NSSI among young people in the general community using data from two large surveys of high school and college students. Two research questions guided the analysis: (1) What factors are most strongly associated with current NSSI, compared with lifetime NSSI? and (2) What factors show consonant and distinct relationships with current NSSI for adolescents and young adults?

86

CURRENT VERSUS LIFETIME NSSI AMONG YOUNG PEOPLE METHODS

Study Design and Samples High School Students. Data for high school students came from the 2010 Minnesota Student Survey (MSS), a populationbased survey conducted by the Minnesota Departments of Education, Health, Human Services, and Public Safety. To our knowledge, this represents the largest survey of high school students that includes an item assessing self-injury, as well as items regarding possible correlates of the behavior. The survey is administered in the spring every 3 years to students in grades 6, 9, and 12. In 2010, 88% of all public school districts participated in the survey, including approximately 79% of 6th grade, 75% of 9th grade, and 59% of 12th grade students (N = 130,908). Additional details concerning the survey methodology are available elsewhere (Minnesota Student Survey Interagency Team, 2010). For the current analysis, only students in grades 9 and 12 (n = 84,121) were included because some risk correlate items differed among the 6th graders. The analytic sample comprised 9,985 students who reported NSSI, 11.9% of the total sample of 9th and 12th grade students. Among students categorized into the current or lifetime NSSI groups (described below), 5,731 (57.4%) were in grade 9 and 4,254 (42.6%) were in grade 12. The analytic sample included 7,220 females (72.3%) and 2,765 males (27.7%). Students reported their race/ethnicity as White (73.9%), African/African American (3.0%), Asian/Pacific Islander (5.2%), Hispanic (4.6%), American Indian (1.4%), mixed race (9.9%), and “I don’t know” (2.0%). College Students. Data for college students came from the spring 2010 National College Health Assessment (NCHA) II, conducted by the American College Health Association. The survey is administered twice a year using volunteer samples of college students from numerous colleges and universities across the United

States. The spring 2010 sample comprised 95,712 students from 139 postsecondary institutions. Additional details concerning the survey methodology are available elsewhere (American College Health Association, 2010). For the current analysis, only typical college-aged students (18–22 years) were included (n = 69,246; 74.3%). The analytic sample comprised 7,801 students who reported NSSI, 10.9% of the total sample. Among students categorized into the current or lifetime NSSI groups (described below), 3,255 (42.7%) were aged 18–19 and 4,363 (57.3%) were aged 20–22. The analytic sample included 5,696 females (74.5%) and 1,946 males (25.5%). Students reported their race/ethnicity as White (72.0%), African/African American (2.5%), Asian/Pacific Islander (8.3%), Hispanic (6.3%), American Indian (0.3%), mixed race (9.0%), and other (1.6%). The institutional review board at the University of Missouri approved this secondary data analysis. Measures NSSI groups were created based on students’ responses to items regarding selfinjury and suicide attempt(s). The MSS included the following items: self-injury– “Have you ever hurt yourself on purpose (cutting, burns, bruises)?” and suicide attempt – “Have you ever tried to kill yourself?” Response options for both items were “no,” “yes, during the last year,” and “yes, more than a year ago.” Items in the NCHA were: self-injury – “Have you ever intentionally cut, burned, bruised, or otherwise injured yourself?” and suicide attempt – “Have you ever attempted suicide?” Response options for both items were “no, never,” “no, not in the last 12 months,” “yes, in the last 2 weeks,” “yes, in the last 30 days,” and “yes, in the last 12 months.” We categorized students into mutually exclusive NSSI groups on the basis of reported “current” self-injury behavior (i.e., in the last 12 months) or “lifetime” behavior (i.e., more than 12 months ago). As our self-injury items did not specify behavior

TALIAFERRO

AND

MUEHLENKAMP

without suicidal intent, we needed to ensure we did not confound self-injury with a suicide attempt. Therefore, we imposed strict criteria for inclusion in the groups regarding the latter; that is, no history of a suicide attempt. As many individuals who engage in NSSI also report experiencing suicidal ideation at some point in their lives (Klonsky et al., 2011), we did not exclude those who reported having thoughts about suicide. Thus, the current NSSI groups included students who reported self-injury during the past year and never attempting suicide. The lifetime groups comprised those who reported engaging in self-injury more than a year ago, yet never attempted suicide in their lifetimes. Researchers have used similar items and group categorization in epidemiological studies with community samples of adolescents (Hilt et al., 2008; Taliaferro et al., 2012). Independent variables were selected based on their documented association with suicidality and/or NSSI among young people (e.g., Borowsky, Ireland, & Resnick, 2001; Jacobson & Gould, 2007; Taliaferro et al., 2012) and inclusion in both the MSS and NCHA. Table 1 presents items for each variable. Measures reflected items commonly used in other epidemiological studies of young people (Borowsky et al., 2001; Centers for Disease Control & Prevention, 2012; Taliaferro et al., 2012). Covariates included gender (female vs. male), age (MSS: grade 9 vs. grade 12; NCHA: 18–19 vs. 20–22), and race/ethnicity (White vs. non-White). Data Analysis Analyses were performed in two stages using SAS 9.3 (SAS Institute, Inc., Cary, NC) to determine factors most strongly associated with current versus lifetime NSSI. Chi-square tests were used to examine relationships between NSSI and each independent variable one at a time for high school and college students. Significant variables were entered simultaneously into

87 logistic regression models to determine factors that distinguished the current NSSI groups from the lifetime groups. Demographic variables were included in all regression models. Effect sizes (Cohen’s d) were calculated to further assess the impact of specific variables.

RESULTS

High School Student Sample Overall, 6.3% of high school students reported current NSSI, and 5.6% reported a lifetime history of self-injury (Table 2). Among high school students, overall, females were more likely to report NSSI. However, the difference in prevalence among females who reported current versus lifetime NSSI was slight (8.6% vs. 8.4%). In contrast, among adolescent males, the difference in rates of behavior between the groups was greater, demonstrating a higher prevalence of current than lifetime NSSI (3.9% vs. 2.7%). Other results for high school students regarding demographics showed that younger students (7.9%) and those who identified as non-White (7.1%) were more likely to report current NSSI than older students (4.3%) and those who identified as White (6.0%). The findings from bivariate tests comparing students who currently engaged in NSSI with those who previously engaged in the behavior are presented in Table 3. Among adolescents, current self-injurers were more likely than youth with a history of NSSI to report depressive symptoms (47.3% vs. 27.4%), hopelessness (36.4% vs. 18.2%), anxiety (33.5% vs. 20.3%), verbal or physical assault victimization (56.0% vs. 42.2%), greater stress (58.0% vs. 51.7%), negative perceptions of their weight (46.3% vs. 41.1%), and a same-sex sexual experience (7.1% vs. 5.6%). Also, youth currently engaged in NSSI were less likely than those with a history of NSSI to report tobacco (22.6% vs. 24.9%) and marijuana use (34.1% vs. 37.4%).

88

CURRENT VERSUS LIFETIME NSSI AMONG YOUNG PEOPLE

TABLE 1

Independent Variables Variable Depressive symptoms

Hopelessness

Stress

Anxiety

Negative perception of weight

Tobacco use

Alcohol use

Binge drinking

Marijuana use

Description of Variable MSS: Often unhappy, depressed, or tearful (dichotomized to agree or mostly agree/mostly disagree or disagree) NCHA: Ever felt so depressed that it was difficult to function (dichotomized to yes, in the last 2 weeks or last 30 days/never, not in the last 12 months; yes, in the last 12 months) MSS: During the last 30 days, felt so discouraged or hopeless wondered whether anything was worthwhile (dichotomized to extremely so or quite a bit/some, a little, or not at all) NCHA: Ever felt things were hopeless (dichotomized to yes, in the last 2 weeks or last 30 days/never, not in the last 12 months; yes, in the last 12 months) MSS: During the last 30 days, felt were under stress or pressure (dichotomized to almost more than could take or quite a bit of pressure/more than usual, a little, or no) NCHA: Ever felt overwhelmed by all had to do (dichotomized to yes, in the last 2 weeks or last 30 days/never, not in the last 12 months; yes, in the last 12 months) MSS: During the last 30 days, felt nervous, worried, or upset (dichotomized to all or most of the time/some, a little, or none of the time) NCHA: Ever felt overwhelming anxiety (dichotomized to yes, in the last 2 weeks or last 30 days/never, not in the last 12 months; yes, in the last 12 months) MSS: At the present time, think you are underweight, about the right weight, or overweight (dichotomized to underweight or overweight/about the right weight) NCHA: How do you describe your weight (dichotomized to very underweight or very overweight/slightly underweight, about the right weight, or slightly overweight) MSS: During the last 30 days, smoked cigarettes, smoked cigars, or used smokeless tobacco (dichotomized to 3 or more days/two days or less) NCHA: Within the last 30 days, used cigarettes, cigars, or smokeless tobacco (dichotomized to 3 or more days/2 days or less) MSS: During the last 30 days, drank one or more drinks of an alcoholic beverage (dichotomized to 3 or more days/2 days or less) NCHA: Within the last 30 days, used alcohol (dichotomized to 3 or more days/2 days or less) MSS: Over the last 2 weeks, times had 5 or more drinks in a row (dichotomized to once or more/never) NCHA: Over the last 2 weeks, times had 5 or more drinks in a row (dichotomized to once or more/never) MSS: During the last 12 months, occasions used marijuana or hashish (dichotomized to 1 or more times/0 times) NCHA: Within the last 30 days, days used marijuana (dichotomized to 1 or more days/not in the last 30 days or never) (continued)

TALIAFERRO

AND

MUEHLENKAMP

89

TABLE 1

(continued) Variable Prescription drug misuse

Multiple sexual partners

Same-sex sexual experience

Victim of verbal or physical assault

Involvement in dating violence

Description of Variable MSS: During the last 12 months, occasions used ADHD or ADD drugs, pain relievers, or tranquilizers/sedatives/barbiturates not prescribed for you by a doctor to get high (dichotomized to 1 or more times/0 times) NCHA: In the last 12 months, taken pain killers, sedatives, or stimulates not prescribed to you (yes/no) MSS: Ever had sexual intercourse; during the last 12 months, with how many different male/female partners had sexual intercourse (dichotomized to 3 or more partners/less than 3 partners) NCHA: Within the last 12 months, with how many partners had oral sex, vaginal intercourse, or anal intercourse (dichotomized to 3 or more partners/less than 3 partners) MSS: During the last 12 months, with how many different male/female partners had sexual intercourse (dichotomized to had sex with someone of the same sex/never had sex or had heterosexual sex) NCHA: Within the last 12 months, had sexual partner(s) who were female/male (yes/no) MSS: During the last 12 months, a student threatened; pushed, shoved, or grabbed; kicked, bitten, or hit you on school property (yes/no) NCHA: Within the last 12 months, were physically assaulted or verbally threatened (yes/no) MSS: Someone were going out with ever hit, hurt, or threatened you; ever forced you to have sex or do something sexual when you didn’t want to (yes/no) NCHA: Within the last 12 months, been in an intimate relationship that was physically or sexually abusive (yes/no)

Note. MSS = Minnesota Student Survey; NCHA = National College Health Assessment; ADHD = attention deficit hyperactivity disorder; ADD = attention deficit disorder

Findings from logistic regression analyses (Table 4) indicated that factors associated with current NSSI among high school students included younger age (d = .96), greater depressive symptoms (d = .41), more hopelessness (d = .40), male gender (d = .27), greater anxiety (d = .23), and being the victim of a verbal or physical assault (d = .14). College Student Sample Among the college student sample, 3.6% of students reported current NSSI, and 7.3% reported a lifetime history of NSSI (Table 2). Similar to high school students, overall, female college students were more likely than males to report NSSI (12.5% vs. 8.1%). Yet the difference in prevalence only

appeared significant for lifetime behavior (F: 8.7%; M: 5.0%). Marked differences on the other demographic characteristics between groups also only appeared for lifetime NSSI, with older students (7.7%) and those who identified as White (7.7%) being more likely to report a history of, but no current engagement in, NSSI compared with younger students (7.0%) and those who identified as non-White (6.6%). Consistent with the high school sample, factors associated with current selfinjury, compared to a history of the behavior, among college students included depressive symptoms (41.7% vs. 24.5%), hopelessness (51.9% vs. 34.9%), anxiety (52.1% vs. 42.8%), being the victim of a verbal or physical assault (35.5% vs.

5.6 (4,694) 2.7 (1,120) 8.4 (3,574) 4.2 (2,000) 7.3 (2,694) 5.6 (3,534) 5.6 (1,138)

6.3 (5,291)

3.9 (1,645) 8.6 (3,646)

7.9 (3,731) 4.3 (1,560)

6.0 (3,800) 7.1 (1,459)

Lifetime NSSI

88.4 (55,675) 87.3 (17,907)

87.9 (41,656) 88.4 (32,480)

93.4 (39,013) 83.0 (35,123)

88.1 (74,136)

Not in Either Group

3.1 (746) 3.8 (1,738) 3.8 (1,147) 3.4 (1,330) 3.6 (1,738) 3.7 (767)

792.8, p < .001 30.9, p < .001

3.6 (2,554)

Current NSSI

2240.1, p < .001

Chi-square

Percentage (frequency)

7.7 (3,788) 6.6 (1,379)

7.0 (2,108) 7.7 (3,033)

5.0 (1,200) 8.7 (3,958)

7.3 (5,247)

Lifetime NSSI

88.7 (48,966) 89.7 (18,767)

89.2 (29,994) 88.9 (34,889)

92.0 (22,258) 87.5 (45,516)

89.1 (71,761)

Not in Either Group

College Student Sample

28.1, p < .001

20.3, p < .001

358.0, p < .001

Chi-square

For high school students reflects 9th grade vs. 12th grade, and for college students reflects students aged 18–19 vs. those aged 20–22.

a

Total Gender Male Female Agea Young Older Race White Non-White

Current NSSI

High School Student Sample

Prevalence of NonSuicidal Self-Injury (NSSI)

TABLE 2

90 CURRENT VERSUS LIFETIME NSSI AMONG YOUNG PEOPLE

TALIAFERRO

AND

MUEHLENKAMP

91

TABLE 3

Percentages, Frequencies, and Effect Sizes for Risk Factors Associated with Current Versus Lifetime NSSI High School Student Sample

Variables Depressive symptoms Hopelessness Anxiety Victim of verbal/physical assault Stress Negative perception of weight Same-sex sexual experience Tobacco use Marijuana use Involvement in dating violence Binge drinking Alcohol use Multiple sexual partners Prescription drug misuse

College Student Sample

Current NSSI (n = 5,291)

Lifetime NSSI (n = 4,694)

Current NSSI (n = 2,554)

Lifetime NSSI (n = 5,247)

% (n)

Chi-square φ

% (n)

Chi-square φ

47.3 (2,465) 27.4 (1,271) 412.5***

.20 41.7 (1,062) 24.5 (1,281) 242.5***

.18

36.4 (1,911) 18.2 (850) 408.0*** 33.5 (1,763) 20.3 (951) 217.1*** 56.0 (2,957) 42.2 (1,974) 187.8***

.20 51.9 (1,317) 34.9 (1,819) 205.1*** .15 52.1 (1,322) 42.8 (2,235) 59.4*** .14 35.5 (904) 28.8 (1,510) 35.5***

.16 .09 .07

58.0 (3,046) 51.7 (2,420) 46.3 (2,419) 41.1 (1,912)

.06 76.1 (1,935) 76.8 (4,023) .05 6.6 (169) 4.9 (258)

0.7 9.5**

.01 .04

8.0**

.03

6.2*

.03

6.7** 11.1*** 2.3

.03 .03 .02

7.6** 2.6 21.5***

.03 .02 .05

7.1 (305)

5.6 (222)

22.6 (1,176) 24.9 (1,154) 34.1 (1,714) 37.4 (1,693) 26.4 (1,393) 27.8 (1,302) 23.1 (1,151) 24.6 (1,110) 20.1 (1,006) 21.5 (970) 30.1 (636) 27.9 (676) 13.3 (643)

12.7 (555)

38.8*** 27.0***

8.4 (213)

6.8 (356)

20.8 (530) 18.2 (952) 27.5 (698) 25.8 (1,343) 7.3 (187) 4.8 (249)

3.3 2.7 2.7

.02 42.4 (1,078) 40.0 (2,088) .02 53.0 (1,342) 52.9 (2,754) .02 23.1 (586) 22.0 (1,146)

4.3* 0.1 1.3

.02 .00 .01

0.8

.01

1.4

.01

23.6 (601) 22.4 (1,169)

Note. Effect sizes reflect Phi coefficients (small = .10, medium = .30, large = .50). *p < .05, **p < .01, ***p < .001.

28.8%), negative perceptions of one’s weight (6.6% vs. 4.9%), and a same-sex sexual experience (8.4% vs. 6.8%) (Table 3). Additional factors associated with current versus lifetime NSSI in this sample were greater tobacco use (20.8% vs. 18.2%), involvement in dating violence (7.3% vs. 4.8%), and binge drinking (42.2% vs. 40.0%). In the final analysis, factors that distinguished current and lifetime college student self-injurers were greater depressive symptoms (d = .44), more hopelessness (d = .35), male gender (d = .27), involvement in dating violence (d = .25), negative

perceptions of one’s weight (d = .20), a same-sex sexual experience (d = .18), younger age (d = .14), being the victim of a verbal or physical assault (d = .11), and identifying as non-White (d = .11) (Table 4).

DISCUSSION

In this study we sought to address gaps in research on NSSI regarding factors associated with current versus lifetime behavior among adolescents and young adults. Findings can inform clinical practice,

92

CURRENT VERSUS LIFETIME NSSI AMONG YOUNG PEOPLE

TABLE 4

Odds Ratios (OR), 95% Confidence Intervals (95% CI), and Effect Sizes for Factors Associated with Current Versus Lifetime NSSI High School Student Sample OR (95% CI) Young age vs. older age Depressive symptoms Hopelessness Female vs. male Involvement in dating violence Anxiety Victim of verbal or physical assault Same-sex sexual experience Negative perception of weight White vs. non-White Binge drinking Tobacco use Stress Marijuana use

College Student Sample d

3.23 1.68 1.66 0.71

(2.91,3.57)*** (1.50,1.89)*** (1.46,1.89)*** (0.62,0.81)***

.96 .41 .40 .27

1.34 1.19 1.16 1.09 1.07

(1.18,1.52)*** (1.08,1.31)*** (0.94,1.43) (0.99,1.21) (0.96,1.20)

.23 .14 .12 .07 .05

0.98 (0.85,1.12) 1.03 (0.92,1.14) 1.01 (0.89,1.13)

.02 .02 .01

OR (95% CI) 1.24 1.74 1.55 0.71 1.37 0.93 1.15 1.25 1.29 0.87 1.08 1.05

(1.12,1.37)*** (1.52,1.99)*** (1.37,1.76)*** (0.63,0.80)*** (1.11,1.70)** (0.83,1.06) (1.03,1.29)* (1.03,1.50)* (1.04,1.60)* (0.78,0.97)* (0.97,1.20) (0.92,1.20)

d .17 .44 .35 .27 .25 .06 .11 .18 .20 .11 .06 .04

Note. All variables were entered simultaneously. Missing primary independent variables were not significant in preliminary tests. Effect sizes reflect Cohen’s d (small = .20, medium = .50, large = .80). *p < .05, **p < .01, ***p < .001.

public health interventions, and future research. Among both samples of high school and college students, females were more likely than males to report NSSI. These findings support previous research (Cloutier, Martin, Kennedy, Nixon, & Muehlenkamp, 2010; Kaminski et al., 2010; Taliaferro et al., 2012). However, a closer examination of differences revealed that among high school students, the difference between current and lifetime NSSI was much greater in males. Females reported almost equivalent rates within both groups, suggesting stability in the rate of NSSI among adolescent females. In contrast, adolescent males appear to currently engage in NSSI more than report a past history of the behavior. This gender difference was not as pronounced among college students, where a difference in prevalence of NSSI only appeared for lifetime behavior, with females reporting a higher rate than males. However, in the multivariate analysis, females within both samples were significantly less

likely than males to report current versus lifetime behavior. This finding supports the notion that females are more likely to engage in NSSI overall, during different time periods of adolescence, and to have a history of NSSI relative to males. However, identifying as male was more strongly associated with a recent act of NSSI relative to females across samples, which may suggest males initiate NSSI at a later age, that an increasing trend exists in the number of males engaging in NSSI, or males engage in the behavior over a longer time period than do females. While the current data cannot address these possibilities, Whitlock and colleagues (2011) reported significant gender differences in practices of NSSI, with males more likely to engage in episodic acts. Episodic NSSI could increase the likelihood of having at least one act in the past year, compared with individuals (predominantly female) who reported more consistent engagement in NSSI that once stopped, may cease to re-occur during the

TALIAFERRO

AND

MUEHLENKAMP

time period assessed. Additional research examining possible gender differences in the trajectory and practice of NSSI across adolescence is strongly encouraged. Overall, our findings support research that suggests NSSI is not solely a female problem (Gollust, Eisenberg, & Golberstein, 2008; Klonsky et al., 2011; Serras, Saules, Cranford, & Eisenberg, 2010), yet NSSI practices may differ for males and females, warranting further study. We also found that younger students were more likely to report NSSI than older students. This finding was especially true for high school students, among whom young age showed the strongest relationship with current NSSI. Findings regarding age support research demonstrating a bimodal distribution of initiation prevalence between ages 13 and 15 (Lofthouse & Yager-Schweller, 2009) and ages 17 and 24 (Whitlock et al., 2006). Thus, clinical practice and prevention programming efforts might target younger students during stressful transition periods in their lives when they may lose important support systems and consider initiating or continuing self-injurious behavior. The transitions to high school and college can involve intense, negative emotions that some youth may have difficulty regulating, resulting in the use of NSSI as a coping mechanism (e.g., Nock, Prinstein, & Sterba, 2009). Few major differences emerged regarding race/ethnicity. In the multivariate analysis, race/ethnicity represented a significant factor to differentiate current and lifetime college student self-injurers, yet the effect was very small. Conflicting evidence exists regarding the association between race/ethnicity and NSSI (Muehlenkamp & Gutierrez, 2007; Taliaferro et al., 2012; Whitlock et al., 2006); thus, more research is needed to draw definitive conclusions. Findings regarding risk factors suggest current NSSI is linked to aversive emotional experiences tied to mood disorder symptoms. Intrapersonal conditions, such as depressive symptoms, hopelessness, anxiety, and negative perceptions of weight, represented especially important risk factors associated with

93 current NSSI. These findings support previous research that suggests emotional and more transient psychological distress symptoms are strongly linked to current engagement in NSSI (Nock et al., 2009). Furthermore, these findings provide additional, albeit indirect, evidence for the selfregulating (intrapersonal) functional model underlying NSSI behavior (Nock & Cha, 2009). Such a model suggests youth engage in NSSI to regulate aversive internal states (emotional or cognitive), representing a maladaptive coping strategy. Findings from this study also suggest the primary psychological symptoms reported as risk factors for NSSI do not differ for adolescents and young adults. Among college students, some additional stressors related to interpersonal conflicts or violence, discussed next, also were associated with current engagement in NSSI; whereas a general sense of stress/pressure was more strongly associated with current self-injury among high school students than college students. This may suggest college students experience a greater number or variety of events that contribute to distress than do high school students who engage in current NSSI, or that young adults might have a higher threshold for tolerating stress before they engage in NSSI. These findings suggest clinical practice and public health programming consider addressing intrapersonal factors that increase risk of NSSI. Clinicians assessing depressive symptoms and hopelessness among adolescents and young adults are strongly encouraged to include assessment of NSSI among youth presenting with such symptoms. In addition, when working with high school-aged youth, clinicians might consider an adolescent’s level of anxiety, as well as possibly feelings of general stress/ pressure. In clinical encounters with college students, practitioners should evaluate the young person’s attitudes toward his/ her body, sexual orientation/same-sex sexual experiences, involvement in dating violence, and possibly tobacco use and binge drinking because these factors showed differential associations with current NSSI by age group.

94

CURRENT VERSUS LIFETIME NSSI AMONG YOUNG PEOPLE

As indicated earlier, interpersonal/ social elements also were important to understanding risk for current NSSI over lifetime behavior. Given the nature of the conflicts examined (being the victim of a verbal/physical assault and involvement in dating violence), the interpersonal conflicts might contribute directly to NSSI and/or intensify the depression, hopelessness, and anxiety symptoms linked to NSSI. Interpersonal events may contribute to the intense negative emotions youth experience, which then trigger NSSI as a coping mechanism. This interpretation is consistent with findings from Muehlenkamp, Brausch, Quigley, and Whitlock (2013), who found interpersonal events were more strongly associated with initiating NSSI than maintaining the behavior. Conversely, interpersonal violence may lead directly to NSSI. Research has demonstrated significant relationships between violence and NSSI, after accounting for intrapersonal factors such as anxiety and depression (e.g., Levesque, Lafontaine, Bureau, Cloutier, & Dandurand, 2010; Taliaferro et al., 2012). More research examining relationships between different types of interpersonal stressors and NSSI would help clarify the mechanisms that account for these relationships. In the meantime, ensuring young people possess the skills to cope with threatening interpersonal conflicts by means other than selfinjury is important. Adults also must address factors in young people’s environments that might unwittingly support negative peer interactions by, for example, implementing and enforcing comprehensive bullying prevention programs. For college students in particular, involvement in dating violence represented another interpersonal risk factor associated with current NSSI, suggesting wellness resource centers and health care practitioners on college campuses ensure prevention programming is implemented that addresses dating violence among their student body. Although evidence of effective programs remains limited, university personnel could apply interventions that have demonstrated success in

reducing other problem behaviors to dating violence prevention such as motivational interventions, dialectical behavior therapy, mindfulness, and bystander interventions (Shorey et al., 2012). Furthermore, clinicians might consider assessing for NSSI among youth who report involvement in a violent situation, as well as addressing victimization and possibly perpetration among those who present with self-injurious behavior. Finally, understanding the importance of a same-sex sexual experience in increasing risk of engaging in NSSI, especially among college students, is essential. A large body of evidence supports associations between sexual minority status, and NSSI and suicidality among young people (e.g., Sornberger, Smith, Toste, & Heath, 2013; Whitlock et al., 2006, 2011). If they do not already, schools should address sexualityrelated bullying and violence among students, and ensure youth within sexual minority groups have needed support systems to effectively cope with negative intrapersonal and interpersonal experiences. Findings from this study must be interpreted within the context of several strengths and limitations. Our data were obtained through self-report measures and originated from cross-sectional surveys, so analyses could not determine causal relationships. Another limitation involves the measures of NSSI, which lacked specificity. For example, the surveys did not glean information regarding the ages of onset and cessation of NSSI, which could inform conceptualizations of current and lifetime behavior. Also, we needed to use relatively strict criteria for inclusion in the NSSI groups by ensuring students coded into a group had never attempted suicide because the self-injury items did not specify that the type of self-injury assessed was completed without suicidal intent, consistent with the formal definition of NSSI. Research suggests around 10% to 30% of young people from the general community who engage in NSSI have attempted suicide at some point in their lives (Brausch & Gutierrez,

TALIAFERRO

AND

MUEHLENKAMP

2010; Lloyd-Richardson, Perrine, Dierker, & Kelley, 2007; Muehlenkamp & Gutierrez, 2007; Whitlock & Knox, 2007). Thus, our groups likely provided conservative estimates and underrepresented the prevalence of self-injury among young people in the United States, which is preferable to overestimating prevalence rates using an item that lacked specificity regarding the intent of the behavior. Strengths of the study included the size and diversity of the samples. The sample sizes provided sufficient power to examine factors that differentiated young people who engaged in NSSI in the last year from those with a history of the behavior, while excluding those who had attempted suicide. Another strength involved the comparison of numerous risk factors and health-risk behaviors within two samples representing developmental periods in young people’s lives. To accurately examine the most important psychosocial risk factors associated with NSSI, investigators should assess current behavior. Continuing the norm of assessing lifetime history of NSSI might result in overlooking the importance of certain risk factors in explaining current behavior. Furthermore, when conducting assessments and attempting to extrapolate findings from one adolescent age cohort to another, clinicians and researchers should appreciate factors that showed similar associations with current behavior among high school and college students (i.e., gender, age, depression, hopelessness, and assault history), as well as those that demonstrated differential effects by age group (i.e., anxiety, race/ethnicity, negative perceptions

95 of weight, a same-sex sexual experience, and dating violence). Future research could build on these findings by examining possible differences in protective factors associated with current versus lifetime NSSI for adolescents and young adults. Consensus on definitions of current and lifetime behavior would facilitate these efforts. Though most researchers apply similar definitions to current and lifetime behavior as used in this study, these definitions do not fully capture nuances of these groups. For example, differences between the NSSI groups as conceptualized in this study could relate to later initiation of NSSI in the current group, earlier cessation in the lifetime group, or unexamined cohort-related differences based on participants’ ages. Future research examining more specific aspects of NSSI behavior among youth who self-injure within a brief time period around an assessment of the behavior and those with a past history of NSSI would further understanding of nuanced differences between these groups. We encourage investigators to continue examining similarities and differences between these groups using representative samples of young people across the United States. Incorporating a specific item about NSSI into national surveys of youth will facilitate these efforts. Last, prospective research studying the onset, cessation, and patterns of NSSI across genders and developmental periods will advance current understanding of NSSI, as well as potentially identify meaningful subgroups of individuals who selfinjure, which could have important implications for clinical practice and prevention programming.

REFERENCES ALFONSO, M., & DEDRICK, R. (2010). Selfinjury among early adolescents. American Journal of Health Education, 41, 74–84. American College Health Association. (2010). American College Health Association-National College Health Assessment II: Reference group data report spring 2010. Linthicum, MD: Author.

ANDERSON, N., & CROWTHER, J. (2012). Using the experiential avoidance model of nonsuicidal self-injury: Understanding who stops and continues. Archives of Suicide Research, 16, 124–134. ANDREWS, T., MARTIN, G., HASKING, P., & PAGE, A. (2013). Predictors of continuation and cessation of nonsuicidal self-injury. Journal of Adolescent Health, 53, 40–46.

96

CURRENT VERSUS LIFETIME NSSI AMONG YOUNG PEOPLE

BOROWSKY, I., IRELAND, M., & RESNICK, M. (2001). Adolescent suicide attempts: Risks and protectors. Pediatrics, 107, 485–493. BRAUSCH, A., & Gutierrez, P. (2010). Differences in non-suicidal self-injury and suicide attempts in adolescents. Journal of Youth and Adolescence, 39, 233–242. BROWN, S. A., WILLIAMS, K., & COLLINS, A. (2007). Past and recent deliberate self-harm: Emotion and coping strategy differences. Journal of Clinical Psychology, 63, 791–803. Centers for Disease Control and Prevention. (2012). Youth risk behavior surveillance – United States, 2011. MMWR Surveillance Summaries, 61, 1–168. CLOUTIER, P., MARTIN, J., KENNEDY, A., NIXON, M., & MUEHLENKAMP, J. (2010). Characteristics and co-occurrence of adolescent nonsuicidal self-injury and suicidal behaviours in pediatric emergency crisis services. Journal of Youth and Adolescence, 39, 259–269. GOLLUST, S., EISENBERG, D., & GOLBERSTEIN, E. (2008). Prevalence and correlates of self-injury among university students. Journal of American College Health, 56, 491–498. HEATH, N., TOSTE, J., NEDECHEVA, T., & CHARLEBOIS, A. (2008). An examination of nonsuicidal self-injury among college students. Journal of Mental Health Counseling, 30, 137–156. HILT, L. M., NOCK, M. K., LLOYD-RICHARDSON, E. E., & PRINSTEIN, M. J. (2008). Longitudinal study of nonsuicidal self-injury among young adolescents: Rates, correlates, and preliminary test of an interpersonal model. Journal of Early Adolescence, 28, 455–469. JACOBSON, C., & GOULD, M. (2007). The epidemiology and phenomenology of non-suicidal self-injurious behavior among adolescents: A critical review of the literature. Archives of Suicide Research, 11, 129–147. JOINER, T., RIBEIRO, J., & SILVA, C. (2012). Nonsuicidal self-injury, suicidal behavior, and their co-occurrence as viewed through the lens of the interpersonal theory of suicide. Current Directions in Psychological Science, 21, 342– 347. KAMINSKI, J., PUDDY, R., HALL, D., CASHMAN, S., CROSBY, A., & ORTEGA, L. (2010). The relative influence of different domains of social connectedness on self-directed violence in adolescence. Journal of Youth and Adolescence, 39, 460–473. KLONSKY, E. D., MUEHLENKAMP, J., LEWIS, S., & WALSH, B. (2011). Nonsuicidal selfinjury. Cambridge, MA: Hogrefe Publishing. LEVESQUE, C., LAFONTAINE, M., BUREAU, K., CLOUTIER, P., & DANDURAND, C. (2010). The influence of romantic attachment and intimate partner violence on non-suicidal self-injuiry

in young adults. Journal of Youth and Adolescence, 39, 474–483. LLOYD-RICHARDSON, E., PERRINE, N., DIERKER, L., & KELLEY, M. (2007). Characteristics and functions of non-suicidal self-injury in a community sample of adolescents. Psychological Medicine, 37, 1183–1192. LOFTHOUSE, N., & YAGER-SCHWELLER, J. (2009). Nonsuicidal self-injury and suicide risk among adolescents. Current Opinions in Pediatrics, 21, 641–645. Minnesota Student Survey Interagency Team (2010). Minnesota student survey 2010. Roseville, MN: Minnesota Department of Education. MUEHLENKAMP, J., BRAUSCH, A., QUIGLEY, K., & WHITLOCK, J. (2013). Interpersonal features and functions of nonsuicidal self-injury. Suicide and Life-Threatening Behavior, 43, 67–80. MUEHLENKAMP, J., CLAES, L., HAVERTAPE, L., & PLENER, P. (2012). International prevalence of adolescent non-suicidal self-injury and deliberate self-harm. Child and Adolescent Psychiatry and Mental Health, 6, 1–9. MUEHLENKAMP, J., & GUTIERREZ, P. M. (2007). Risk for suicide attempts among adolescents who engage in non-suicidal self-injury. Archives of Suicide Research, 11, 1–14. NOCK, M., & CHA, C. (2009). Psychological models of nonsuicidal self-injury. In M. K. Nock (Ed.), Understanding nonsuicidal self-injury: Origins, assessment, and treatment (pp. 65–77). Washington, DC: American Psychological Association. NOCK, M., PRINSTEIN, M., & STERBA, S. (2009). Revealing the form and function of selfinjurious thoughts and behaviors: A real-time ecological assessment study among adolescents and young adults. Journal of Abnormal Psychology, 118, 816–827. ROTOLONE, C., & MARTIN, G. (2012). Giving up self-injury: A comparison of everyday social and personal resources in past versus current self-injurers. Archives of Suicide Research, 16, 147–158. SERRAS, A., SAULES, K., CRANFORD, J., & EISENBERG, D. (2010). Self-injury, substance use, and associated risk factors in a multi-campus probability sample of college students. Psychology of Addictive Behavior, 24, 119–128. SHOREY, R., ZUCOSKY, H., BRASFIELD, H., FEBRES, J., CORNELIUS, T., SAGE, C., ET AL. (2012). Dating violence prevention programming: Directions for future interventions. Aggression and Violent Behavior, 17, 289–296. SORNBERGER, M. J., SMITH, N. G., TOSTE, J. R., & HEATH, N. L. (2013). Nonsuicidal selfinjury, coping strategies, and sexual orientation. Journal of Clinical Psychology, 69, 571–583.

TALIAFERRO

AND

MUEHLENKAMP

TALIAFERRO, L., MUEHLENKAMP, J., BOROWSKY, I., MCMORRIS, B., & KUGLER, K. (2012). Factors distinguishing youth who report self-injurious behavior: A population-based sample. Academic Pediatrics, 12, 205–213. TAYLOR, J., PETERSON, C., & FISCHER, S. (2012). Motivations for self-injury, affect, and impulsivity: A comparison of individuals with current self-injury to individuals with a history of self-injury. Suicide and Life-Threatening Behavior, 42, 602–613. WHITLOCK, J., ECKENRODE, J., & SILVERMAN, D. (2006). Self-injurious behaviors in a college population. Pediatrics, 117, 1939–1948. WHITLOCK, J., & KNOX, K. (2007). The relationship between self-injurious behavior and suicide in a young adult population.

97 Archives of Pediatric and Adolescent Medicine, 161, 634–640. WHITLOCK, J., MUEHLENKAMP, J., ECKENRODE, J., PURINGTON, A., ABRAMS, G. B., BARREIRA, P., ET AL. (2013). Nonsuicidal self-injury as a gateway to suicide in young adults. Journal of Adolescent Health, 52, 486–492. WHITLOCK, J., MUEHLENKAMP, J., PURINGTON, A., ECKENRODE, J., BARREIRA, P., ABRAMS, G., ET AL. (2011). Nonsuicidal self-injury in a college population: General trends and sex differences. American Journal of College Health, 59, 691–698. Manuscript Received: January 19, 2014 Revision Accepted: May 7, 2014

Copyright of Suicide & Life-Threatening Behavior is the property of Wiley-Blackwell and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.