Contextual Features and Behavioral Functions of Self-Mutilation
Matthew K. Nock
Mitchell J. Prinstein
Adolescent self-mutilative behavior (SMB) is a pervasive and dangerous problem, yet factors influencing
the performance of SMB are not well understood. The authors examined the contextual features and
behavioral functions of SMB in a sample of 89 adolescent psychiatric inpatients. SMB typically was
performed impulsively, in the absence of physical pain, and without the use of alcohol or drugs.
Moreover, analyses supported the construct validity of a functional model in which adolescents reported
engaging in SMB for both automatic and social reinforcement. Considering the functions of SMB
clarified the relations between SMB and other clinical constructs reported in previous studies such as
suicide attempts, posttraumatic stress, and social concerns and has direct implications for the assessment
and treatment of SMB.
Self-mutilative behavior (SMB) refers to the direct and deliber-
ate destruction of one’s own body tissue without suicidal intent.
SMB is a pervasive public health problem occurring at a rate of 4%
in the general adult population and 21% in adult clinical popula-
tions (Briere & Gil, 1998; Klonsky, Oltmanns, & Turkheimer,
2003). Adolescence is a period of significantly increased risk for
SMB, as is evidenced by rates of 14%–39% in adolescent com-
munity samples (Lloyd, Kelley, & Hope, 1997; Ross & Heath,
2002) and 40%–61% in adolescent psychiatric inpatient samples
(Darche, 1990; DiClemente, Ponton, & Hartley, 1991).
Despite the alarming prevalence and obvious dangerousness of
SMB, it has received relatively little research attention and remains
a poorly understood behavioral phenomenon. Several factors have
contributed to the slow progress of research in this area. First,
practical constraints, including the successful recruitment of re-
search participants, the collection of reliable data, and the ethical
and legal deterrents involved in the study of self-harm, have been
a primary impediment to controlled research on SMB. Second,
most previous studies of SMB have not used direct, systematic
evaluation of self-mutilators but have taken the form of clinical
case reports (e.g., Noshpitz, 1994; Offer & Barglow, 1960), large
survey studies (e.g., Favazza & Conterio, 1989; Ross & Heath,
2002), or laboratory studies with healthy, nonmutilating partici-
pants (e.g., McCloskey & Berman, 2003). Variations in sampling
and measurement methods used across these different studies have
led to inconsistent and sometimes contradictory findings regarding
the nature of SMB. Third, the few studies that have incorporated
direct, systematic evaluation of self-mutilators generally have
failed to move beyond evaluations of the clinical correlates of
SMB and have neither examined the context in which SMB occurs
nor tested theories about factors initiating or maintaining SMB.
For instance, previous studies of adolescent SMB indicate it is
correlated with a range of clinical constructs including: previous
suicide attempts, hopelessness, symptoms of depression and anx-
iety, past loss or abuse, perfectionism, and loneliness, to name a
few (Darche, 1990; Guertin, Lloyd-Richardson, Spirito, Donald-
son, & Boergers, 2001; Klonsky et al., 2003; Penn, Esposito,
Schaeffer, Fritz, & Spirito, 2003; Walsh & Rosen, 1988). Although
the identification of such correlates may help predict who is at risk
for SMB, this strategy does little to advance the understanding of
why individuals engage in SMB—this is the primary goal of the
There are many reasons individuals might engage in SMB.
Some theoretical models suggest SMB is performed for the pur-
pose of boundary definition, mastery of penetration and other
sexual impulses, or mastery over death (see Suyemoto, 1998);
however, perhaps as a result of problems operationalizing and
measuring such constructs, these models have not received empir-
ical support. More systematic research has demonstrated SMB
may serve multiple functions such as the reduction of tension or to
communicate with others (Brown, Comtois, & Linehan, 2002;
Haines, Williams, Brain, & Wilson, 1995) Similar multifunction
models of self-injury have been hypothesized and carefully eval-
uated among those with developmental disabilities (see Iwata et
al., 1994) and may prove useful with nondevelopmentally disabled
populations as well.
Building on these earlier findings, we recently developed and
evaluated a comprehensive theoretical model that proposes four
primary functions of SMB that differ along two dichotomous
dimensions: contingencies for SMB that are automatic (i.e., within
Matthew K. Nock, Department of Psychology, Havard University;
Mitchell J. Prinstein, Department of Psychology, Yale University.
Support for this work was provided in part by National Institute of
Mental Health Grant MH11770 to Mitchell J. Prinstein. We thank Eliza-
beth Lloyd-Richardson for the use of the Functional Assessment of Self-
Mutilation in this study and Dana Damiani and Joy Richmond for their
assistance with data collection.
Correspondence concerning this article should be addressed to Matthew
K. Nock, Department of Psychology, Harvard University, 33 Kirkland
Street, Cambridge, MA 02138 or to Mitchell J. Prinstein, who is now at the
Department of Psychology, University of North Carolina at Chapel Hill,
240 Davie Hall, Chapel Hill, NC 27599-3270. E-mail: firstname.lastname@example.org
Journal of Abnormal Psychology
2005, Vol. 114, No. 1, 140–146
Copyright 2005 by the American Psychological Association
oneself)1versus social, and reinforcement that is positive (i.e.,
followed by the presentation of a favorable stimulus) versus neg-
ative (i.e., followed by the removal of an aversive stimulus; see
Nock & Prinstein, 2004). Confirmatory factor analyses and reli-
ability analyses supported the structural validity and reliability of
this four-function model, with adolescents reporting engagement
in SMB for automatic negative reinforcement (e.g., “To stop bad
feelings”), automatic positive reinforcement (e.g., “To feel some-
thing, even if it is pain”), social negative reinforcement (e.g., “To
avoid doing something unpleasant you do not want to do”) and
social positive reinforcement (e.g., “To get attention”).
The primary goal of the current study was to extend this func-
tional model in two ways. First, given that our four function model
focuses solely on the reinforcement of SMB, our initial goal in the
current study was to examine some of the contextual features that
precede SMB (i.e., behavioral antecedents) as well as factors that
may serve to punish SMB. We focused on several contextual
features that have been discussed in previous reports but whose
actual role in the performance of adolescent SMB has not been
directly and systematically evaluated in a large clinical sample,
including: the impulsiveness of SMB, the use of alcohol and drugs
prior to SMB, the role of social modeling in the initiation of SMB,
and the absence of physical pain experienced as a result of SMB.
The information obtained in this study about the role of each of
these contextual factors in the performance of adolescent SMB
should have significant implications for researchers and clinicians
alike. For instance, if alcohol use and social modeling influence
the performance of self-harm, as suggested in an experimental
study among healthy adult participants (McCloskey & Berman,
2003), then these factors should be targeted in research and inter-
vention efforts. However, if these factors do not play a significant
role in the performance of SMB, as was suggested in a large
mail-in survey study of female adult self-mutilators (Favazza &
Conterio, 1989), then researchers and clinicians may be better
served by focusing on other factors influencing SMB.
Our second goal in this study was to use our functional model of
SMB to help explain the relations between SMB and the hetero-
geneous list of clinical correlates identified in previous studies.
Although it is certainly useful to know what constructs are corre-
lated with SMB, a more detailed understanding of how and why
these constructs are related to SMB would further elucidate the
nature of SMB and facilitate research and clinical work in this
area. We believed a consideration of the functions of SMB would
help explain each of the previously observed correlations. Drawing
from a functional perspective, we expected SMB to correlate with
other clinical constructs for one of two reasons: The construct
represents an antecedent condition or process that increases the
probability of SMB, or the construct represents a functionally
equivalent behavior. On the basis of such a model, the four
functions of SMB should not relate uniformly to each of the
clinical correlates of SMB, but rather each function should be
related only with clinical constructs that represent a related ante-
cedent process/dysfunction or that serve a similar function (con-
vergent validity) and not with each of the other constructs (diver-
Accordingly, on the basis of this functional theoretical model,
we hypothesized a specific pattern of relations between the four
functions of SMB and the associated clinical constructs. First, we
hypothesized hopelessness and previous suicide attempts are
uniquely associated with the automatic negative reinforcement
function of SMB. That is, hopelessness represents a negative
cognitive state from which one is likely to try to escape (i.e., an
antecedent condition increasing likelihood of escape behavior) and
suicide attempts represent a behavioral response that is often
performed for the purposes of escape from negative affective or
cognitive states (i.e., a functionally equivalent behavior; see
Boergers, Spirito, & Donaldson, 1998; Hawton, Cole, O’Grady, &
Osborn, 1982). Conversely, we did not expect engaging in SMB
for automatic positive reinforcement or social communication to
be associated with hopelessness or making a suicide attempt.
Second, we hypothesized that symptoms of major depressive dis-
order (MDD) and posttraumatic stress disorder (PTSD) are
uniquely associated with the automatic positive function, given the
anhedonia, inactivity, and psychic numbness associated with these
diagnoses are most likely to precipitate the need for SMB per-
formed for feeling generation. Third, we hypothesized both social
reinforcement functions of SMB are uniquely associated with
social concerns such as loneliness and socially prescribed (vs.
self-oriented) perfectionism. The demonstration of these specific
relations would support the construct validity of our functional
model of SMB but more importantly would enhance understanding
of this complex behavioral phenomenon and its relations with
other forms of psychopathology.
Participants were 89 (23 boys, 66 girls) adolescent (12–17 years; M ?
14.7, SD ? 1.4) psychiatric inpatients who reported engaging in SMB in
the previous 12 months. The ethnic composition of the sample was 76.4%
European American, 8.9% Latin American, 4.5% African American, and
10.1% Mixed Ethnicity/Other. According to state census tract data, socio-
economic status for adolescents in this sample was: 3.0% High, 57.6%
Moderate, 24.2% Low, and 15.2% Poverty. Participants in this study were
included in our previous study of the functions of SMB (Nock & Prinstein,
2004); however, the current study examines a unique set of hypotheses and
is thus reported separately.
Data were obtained via comprehensive interviews administered on ad-
mission to an adolescent psychiatric inpatient unit. For participants admit-
ted to the unit on more than one occasion during the study period, only data
from the first admission were used. All study procedures were approved by
the Institutional Review Board.
last 12 months was evaluated using the Functional Assessment of Self-
Mutilation (FASM; Lloyd et al., 1997). Adolescents provided information
about the frequency of different methods of SMB (i.e., whether and how
many times they had engaged in behaviors such as cutting, burning, and
inserting objects under their skin) and other characteristics of this behavior,
Participants’ engagement in SMB over the
1We use the term automatic to refer to reinforcement that is conducted
or carried out by oneself (consistent with previous research in applied
behavior analysis) and not to refer to behavior that is performed without
conscious effort (a definition often used in cognitive psychology).
including the amount of time they thought about each incident before
engaging in the behavior, the degree of physical pain experienced, the use
of alcohol and/or drugs during SMB, and their knowledge of the perfor-
mance of SMB by their friends.
Participants also indicated on a scale ranging from 0 (never) to 3 (often)
how often they had engaged in SMB for each of 22 different reasons. As
supported by confirmatory factor analysis and reliability analyses (Nock &
Prinstein, 2004) 21 of the items were placed into one of four subscales:
automatic negative reinforcement (2 items: ? ? .62), automatic positive
reinforcement (3 items; ? ? .69), social negative reinforcement (4 items:
? ? .76), and social positive reinforcement (12 items: ? ? .85). Subscale
item means were used as the measure of each construct. A complete list of
items, factor loadings, and frequency of endorsement of each method and
function of SMB are included in Nock and Prinstein (2004).
Participants’ number of symptoms of MDD,
PTSD, and past suicide attempt status were evaluated with the Diagnostic
Interview Schedule for Children (DISC; Shaffer et al., 1996), a structured
clinical interview developed for use with children and adolescents ages 6
to 17 years. The DISC contains items that assess current and past symp-
toms, behaviors, and emotions corresponding to criteria from the Diagnos-
tic and Statistical Manual of Mental Disorders (4th ed.; DSM–IV; Amer-
ican Psychiatric Association, 1994). The DISC has adequate diagnostic
sensitivity (Fisher et al., 1993), test–retest reliability (Jensen et al., 1996),
and criterion validity (Schwab-Stone et al., 1996). The presence of a
suicide attempt in the previous 4 weeks was assessed via adolescents’
self-report on the question: “In the past 4 weeks, have you tried to kill
yourself?” This is consistent with the common use of structured interview
items for the measurement of suicide-related constructs in adolescents
(Nock & Kazdin, 2002; Prinstein, Nock, Spirito, & Grapentine, 2001).
Additional clinical constructs were examined through the use of well
regarded and psychometrically sound rating scales in each domain. Hope-
lessness was assessed using the Hopelessness Scale for Children (Kazdin,
Rodgers, & Colbus, 1986), a 17-item true–false scale measuring adoles-
cents’ negative expectancies for the future. Loneliness was assessed with
the Revised UCLA Loneliness Scale (Russell, Peplau, & Cutrona, 1980), a
20-item, 4-point scale measuring statements describing the experience of
being lonely and socially isolated. Perfectionism was assessed with the
Child and Adolescent Perfectionism Scale (Hewitt & Flett, 1991), a 22-
item, 5-point scale measuring multiple dimensions of perfectionism, with
separate subscales for self-oriented (i.e., exceedingly high self-imposed
self-standards) and socially prescribed (i.e., beliefs that others maintain
exceedingly high standards for oneself) perfectionism.
Data Analytic Plan
Several data analytic strategies were used to evaluate the study hypoth-
eses. First, the contextual features of SMB were evaluated by examining
the descriptive statistics of adolescents’ responses on the FASM. Second,
the associations among the contextual features as well as their relations
with the four functions of SMB were examined with correlational analyses.
Third, after examining and controlling for the variance explained by
demographic variables, relations between each of the four functions of
SMB and the clinical constructs were evaluated via multiple regression
analyses through the use of simultaneous entry.
Contextual Features of SMB
Most self-mutilators contemplated SMB for a few minutes or
less before performing each incident and reported not using alco-
hol or drugs during incidents of SMB, as is shown in Table 1.
Despite the infrequency of alcohol and drug use during SMB, most
adolescents reported experiencing little or no pain during each
incident. In addition, 82.1% of adolescents reported SMB among
at least one of their friends in the previous 12 months with an
average of 4.11 (SD ? 3.17, Mdn ? 4.00, range ? 0–11) SMB
incidents, which suggests that social modeling may play a role in
the performance of SMB.
We performed additional analyses to examine the relations
between these contextual features and the four functions of SMB.
As is reported in Table 2, the four functions of SMB were signif-
icantly correlated. The shared variance among the subscales (15%–
61%) indicates that although significantly related, the four func-
tions represent distinct constructs. Endorsement of a social
function for SMB was significantly associated with the amount of
time spent contemplating SMB before engaging in this behavior
and endorsement of the social positive reinforcement function was
significantly associated with the number of SMB incidents per-
formed by one’s friends. In addition, endorsement of SMB for
positive reinforcement (regardless of whether contingencies were
automatic or social) was significantly associated with the absence
of alcohol or drug use during SMB.
These relations are further clarified with a consideration of the
associations among the contextual features examined. The expe-
rience of more physical pain during SMB was significantly asso-
ciated with the amount of time spent contemplating SMB before
doing it and the number of SMB incidents performed by one’s
friends. Thus, the more SMB hurts, the longer one thinks about it
before doing it and the more likely one is to have a friend who also
engages in SMB. Surprisingly, the experience of less physical pain
was not related with the concurrent use of alcohol and drugs.
Relations Between the Behavioral Functions and Clinical
Correlates of SMB
The relations between the four functions of SMB and adolescent
age, gender, ethnicity, and socioeconomic status were evaluated in
four simultaneous regression analyses. Entered as a block, these
demographic variables failed to account for significant variance in
scores on the automatic negative reinforcement subscale, F(4,
84) ? 2.19, ns, R2? .09, or on the automatic positive reinforce-
ment subscale, F(4, 84) ? 1.94, ns, R2? .08, of the FASM. In
Descriptive Statistics for Contextual Features of SMB
Not at all
A few seconds
A few minutes
Less than one hour
Less than one day
More than 2 days
Friends’ SMB incidents
Note. N ? 89. SMB ? Self-mutilative behavior.
NOCK AND PRINSTEIN
contrast, these demographic variables explained significant vari-
ance in scores on the social negative reinforcement subscale, F(4,
84) ? 2.93, p ? .05, R2? .12, and on the social positive
reinforcement subscale, F(4, 84) ? 4.07, p ? .05, R2? .12. In
these latter analyses, higher scores on both social negative and
social positive reinforcement were associated with younger age,
? ? ?.22 and ?.26, respectively (ps ? .05), and with ethnic
minority status, ? ? .26 and .26, respectively ( ps ? .05). There-
fore, adolescent age and ethnicity were controlled in subsequent
analyses including the social function subscales.
The relations between the four functions of SMB and the clin-
ical constructs mentioned above were evaluated in four simulta-
neous regression analyses. The clinical constructs examined ex-
reinforcement, F(7, 81) ? 2.08, p ? .05, R2? .20; automatic
positive reinforcement, F(7, 81) ? 5.72, p ? .001, R2? .33; social
negative reinforcement, ?F(7, 79) ? 2.72, p ? .05, ?R2? .18;
and social positive reinforcement, ?F(7, 79) ? 4.22, p ? .001,
?R2? .24, function subscales. Moreover, as shown in Table 3, the
pattern of statistically significant relations supported the study
hypotheses and the construct validity of our model. Specifically,
recent suicide attempt and hopelessness were associated with only
the automatic negative reinforcement function of SMB; depressive
and posttraumatic stress symptoms were significantly associated
with the automatic positive reinforcement function, and socially
prescribed (but not self-oriented) perfectionism was related to the
in the automaticnegative
social reinforcement functions. In addition, depressive symptoms
were also associated with the social reinforcement functions; how-
ever, loneliness was not significantly associated with any of the
This study provided a detailed analysis of the contextual fea-
tures and behavioral functions of adolescent SMB, a dangerous
and alarmingly pervasive problem among this population. Our
results indicate adolescent SMB typically is performed impul-
sively, without the use of alcohol or drugs, and in the absence of
physical pain. These findings highlight salient concerns for re-
searchers and clinicians. From a research perspective, this impul-
siveness suggests that the performance of SMB is influenced by
immediate internal and external contingencies rather than as the
result of long-term decision-making processes and planning. How-
ever, it may be that this finding is specific to a population of
individuals who have already engaged in SMB. The initial act of
SMB may occur nonimpulsively, but subsequent acts may occur
without substantial planning. Examination of antecedents to initial
SMB episodes as compared with factors that serve to maintain or
reinforce ongoing SMB is a high priority for future research.
Nevertheless, from a clinical perspective, this impulsiveness and
lack of physical pain is of high concern as this suggests SMB is
difficult to prevent and treat given the limited time frame for
Correlations Among Contextual Features and Behavioral Functions of SMB
6. Alcohol/drug use
8. Friends’ SMB
reinforcement; SNR ? social negative reinforcement; SPR ? social positive reinforcement.
* p ? .05.** p ? .01. *** p ? .001.
SMB ? self-mutilative behavior; ANR ? automatic negative reinforcement; APR ? automatic positive
Relations Between Clinical Correlates and Behavioral Functions of SMB
B SE B
B SE B
B SE B
reinforcement; SPR ? social positive reinforcement. Boldface values represent relations supporting study hypotheses.
* p ? .05.** p ? .01. *** p ? .001.
SMB ? self-mutilative behavior; ANR ? automatic negative reinforcement; APR ? automatic positive reinforcement; SNR ? social negative
intervention and the lack of naturally occurring aversive
Our findings on the absence of physical pain during SMB are
especially interesting. It has been proposed that SMB may be
reinforced via the release of endogenous endorphins and that this
release also serves to block the experience of pain (e.g., Haines et
al., 1995). This may indeed be true in some cases; however, our
results suggest that there is variability in the experience of pain
during SMB. In this study, the experience of pain was associated
with spending more time contemplating each incident of SMB and
having a friend who also engaged in SMB, suggesting complex
relations among pain and internal and social influences. The fur-
ther explication of the relation between the experience of physical
pain and SMB should figure prominently in the research agenda
for SMB research, given the understanding of such relations holds
promise for improving interventions aimed at blocking the sooth-
ing effects of SMB and increasing the naturally aversive conse-
quences of such behaviors.
A substantial proportion of adolescents reported that their
friends (from outside the hospital setting) had also engaged in
SMB. Friends’ behavior may increase adolescents’ access to SMB
through priming as a potential strategy for achieving automatic and
social contingencies. Results indicated that the number of SMB
incidents among friends was significantly associated with a social
positive reinforcement function of SMB suggesting that some
adolescents may believe that their friends’ behavior was successful
in eliciting specific social behaviors from others in the interper-
sonal context. Similar social influences have also been described in
the literature on suicide contagion effects (Gould, 2001; Joiner,
2003) and represent another important avenue for further
Perhaps most importantly, the current study provided important
information about the specific relations between the functions of
SMB and related constructs, and these relations supported the
construct validity of our functional model of SMB. Consistent with
our predictions, the functions of SMB were differentially related to
the clinical correlates from previous studies of SMB. Automatic
negative reinforcement, the most frequently endorsed function
among adolescents (Nock & Prinstein, 2004) was uniquely asso-
ciated with hopelessness and a history of suicide attempts. Previ-
ous studies have reported associations between SMB and past
suicide attempts and hopelessness (Dulit, Fyer, Leon, Brodsky, &
Frances, 1994; Favazza & Conterio, 1989; Stanley, Gameroff,
Michalsen, & Mann, 2001) but have failed to explain the nature of
this relation. Given that adolescents who attempt suicide most
often report they do so to escape negative experiences (Boergers et
al., 1998; Hawton et al., 1982), our results suggest functional
equivalence between these two forms of self-injurious behaviors in
The observed relations between an automatic positive reinforce-
ment function for SMB and symptoms of MDD and PTSD support
our hypothesis that the experience of feelings of emptiness, de-
tachment, anhedonia, and a restricted range of affect may increase
the likelihood of engaging in SMB for automatic positive rein-
forcement to generate certain sensations or feelings. Although not
evaluated in the current study, we would hypothesize that the link
between the experience of a significant loss or abuse during
childhood and subsequent SMB (van der Kolk, Perry, & Herman,
1991; Zlotnick et al., 1996) may be specific to SMB for automatic
positive reinforcement and that symptoms of depression and/or
posttraumatic stress may mediate this relation. Long-term longi-
tudinal analyses are needed to test this hypothesis; however, our
data are consistent with such a model.
Previous research has consistently demonstrated a relation be-
tween perfectionism and self-injurious thoughts and behaviors (see
Blatt, 1995, for a review). In the current study, the convergent and
divergent validity of the social functions of SMB were supported
via specific relations with socially prescribed perfectionism but not
self-oriented perfectionism. Given both social functions were re-
lated with this dimension of perfectionism, it is possible that
adolescents were attempting to use SMB to solicit assistance from
others (i.e., social positive reinforcement) or to remove the per-
ceived expectations of others (i.e., social negative reinforcement).
In addition, both social functions of SMB were significantly
related with younger age, ethnic minority status, and symptoms of
MDD. This last result supports the notion that socially reinforced
SMB is not synonymous with the absence of psychopathology, as
is sometimes implied in clinical settings. It is possible that either
adolescents experience increased depressive symptoms after en-
gaging in self-harm or that depressive symptoms precede self-
harm, which may serve to alleviate depression via a social mech-
anism. For instance, it has been suggested elsewhere that self-harm
behavior may serve to increase support within individuals’ social
networks (Walker, Joiner, & Rudd, 2001). Consistent with this
notion, we recently found in a prospective study that a subgroup of
adolescents who engage in SMB reported subsequent improve-
ments in the quality of their relationship with parents (Hilt, Borelli,
Nock, & Prinstein, 2004).
Overall, these findings provide clear directions for future re-
search and clinical work on the etiologies, assessment, and treat-
ment of SMB. First, these findings extend our functional model of
SMB and offer additional support for its validity, further highlight-
ing the advantages of the use of a functional approach in the study
of SMB. Future research efforts that use this framework are likely
to further enhance understanding of SMB, and clinicians are urged
to use a functional perspective when assessing SMB in clinical
settings. Clearly, all self-mutilators are not alike and the consid-
eration of the function of this behavior can guide one’s clinical and
research conceptualizations. Second, this model offers clear targets
for assessing antecedents and consequences of SMB in both re-
search and clinical settings. This study demonstrated a small set of
specific relations between the different functions of SMB and
several antecedents and clinical correlates. Additional relations
with other constructs are yet to be examined and can be derived
directly from the model. For instance, we would anticipate auto-
matic negative reinforcement to be related to high emotional
reactivity, automatic positive reinforcement to be related to low
reactivity and dissociation, and the social reinforcement functions
to be associated with additional problems with social interaction
and communication skills. Third, these findings suggest different
treatment approaches may be warranted depending on the function
of SMB. Researchers and clinicians who develop and evaluate
such treatments should consider the different antecedents, corre-
lates, and consequences associated with each function and tailor
treatments accordingly, rather than by using a one-size-fits-all
approach to the treatment of SMB.
Despite the strengths and implications of this study, several
methodological factors limit the inferences that can be drawn.
NOCK AND PRINSTEIN
First, although evaluating the functions of SMB through the use of
self-report allowed us to gather data on multiple SMB incidents
performed in a wide range of settings and for different reasons, the
use of this methodology may suffer from potential biases (e.g.,
social desirability) and inaccuracies (e.g., distortions of retrospec-
tive recall). Second, our sample consisted of adolescent psychiatric
inpatients, thus, these results may not generalize to other age
groups or other settings. Third, these data were cross-sectional,
limiting our ability to make conclusive statements about the direc-
tion of the observed relations among study constructs. Future
studies are planned assessing the antecedents and consequences of
SMB at multiple time points across a wide range of real-world
settings to overcome these limitations and to further explicate the
determinants of adolescent SMB. Continued work in this direction
is likely to further develop our understanding of SMB and is
necessary to improve methods for assessing and treating this
pervasive and dangerous behavior problem.
American Psychiatric Association. (1994). Diagnostic and Statistical Man-
ual of Mental Disorders (4th ed.). Washington, DC: Author.
Blatt, S. J. (1995). The destructiveness of perfectionism. Implications for
the treatment of depression. American Psychologist, 50, 1003–1020.
Boergers, J., Spirito, A., & Donaldson, D. (1998). Reasons for adolescent
suicide attempts: Associations with psychological functioning. Journal
of the American Academy of Child and Adolescent Psychiatry, 37,
Briere, J., & Gil, E. (1998). Self-mutilation in clinical and general popu-
lation samples: Prevalence, correlates, and functions. American Journal
of Orthopsychiatry, 68, 609–620.
Brown, M. Z., Comtois, K. A., & Linehan, M. M. (2002). Reasons for
suicide attempts and nonsuicidal self-injury in women with borderline
personality disorder. Journal of Abnormal Psychology, 111, 198–202.
Darche, M. A. (1990). Psychological factors differentiating self-mutilating
and non-self-mutilating adolescent inpatient females. The Psychiatric
Hospital, 21, 31–35.
DiClemente, R. J., Ponton, L. E., & Hartley, D. (1991). Prevalence and
correlates of cutting behavior: Risk for HIV transmission. Journal of the
American Academy of Child and Adolescent Psychiatry, 30, 735–739.
Dulit, R. A., Fyer, M. R., Leon, A. C., Brodsky, B. S., & Frances, A. J.
(1994). Clinical correlates of self-mutilation in borderline personality
disorder. American Journal of Psychiatry, 151, 1305–1311.
Favazza, A. R., & Conterio, K. (1989). Female habitual self-mutilators.
Acta Psychiatrica Scandanavica, 79, 283–289.
Fisher, P. W., Shaffer, D., Piacentini, J. C., Lapkin, J., Kafantaris, V.,
Leonard, H., et al. (1993). Sensitivity of the Diagnostic Interview Sched-
ule for Children, 2nd ed. (DISC–2.1) for specific diagnoses of children
and adolescents. Journal of the American Academy of Child and Ado-
lescent Psychiatry, 32, 666–673.
Gould, M. S. (2001). Suicide and the media. Annals of the New York
Academy of Sciences, 932, 200–221.
Guertin, T., Lloyd-Richardson, E., Spirito, A., Donaldson, D., & Boergers,
J. (2001). Self-mutilative behavior in adolescents who attempt suicide by
overdose. Journal of the American Academy of Child and Adolescent
Psychiatry, 40, 1062–1069.
Haines, J., Williams, C. L., Brain, K. L., & Wilson, G. V. (1995). The
psychophysiology of self-mutilation. Journal of Abnormal Psychology,
Hawton, K., Cole, D., O’Grady, J., & Osborn, M. (1982). Motivational
aspects of deliberate self-poisoning in adolescents. British Journal of
Psychiatry, 141, 286–291.
Hewitt, P. L., & Flett, G. L. (1991). Dimensions of perfectionism in
unipolar depression. Journal of Abnormal Psychology, 100, 98–101.
Hilt, L. M., Borelli, J. L., Nock, M. K., & Prinstein, M. J. (2004, April).
Interpersonal models of adolescent self-mutilative behavior: A longitu-
dinal study. Paper presented at the National Conference on Child Health
Psychology, Charleston, South Carolina.
Iwata, B. A., Pace, G. M., Dorsey, M. F., Zarcone, J. R., Vollmer, T. R.,
Smith, R. G., et al. (1994). The functions of self-injurious behavior: An
experimental–epidemiological analysis. Journal of Applied Behavior
Analysis, 27, 215–240.
Jensen, P. S., Watanabe, H. K., Richters, J. E., Roper, M., Hibbs, E. D.,
Salzberg, A. D., et al. (1996). Scales, diagnoses, and child psychopa-
thology: II. Comparing the CBCL and the DISC against external vali-
dators. Journal of Abnormal Child Psychology, 24, 151–168.
Joiner, T. (2003). Contagion of suicidal symptoms as a function of assor-
tative relating and shared relationship stress in college roommates.
Journal of Adolescence, 26, 495–504.
Kazdin, A. E., Rodgers, A., & Colbus, D. (1986). The Hopelessness Scale
for Children: Psychometric characteristics and concurrent validity. Jour-
nal of Consulting and Clinical Psychology, 54, 241–245.
Klonsky, E. D., Oltmanns, T. F., & Turkheimer, E. (2003). Deliberate
self-harm in a nonclinical population: Prevalence and psychological
correlates. American Journal of Psychiatry, 160, 1501–1508.
Lloyd, E., Kelley, M. L., & Hope, T. (1997, April). Self-mutilation in a
community sample of adolescents: Descriptive characteristics and pro-
visional prevalence rates. Paper presented at the Annual Meeting of the
Society for Behavioral Medicine, New Orleans, Louisiana.
McCloskey, M. S., & Berman, M. E. (2003). Alcohol intoxication and
self-aggressive behavior. Journal of Abnormal Psychology, 112, 306–
Nock, M. K., & Kazdin, A. E. (2002). Examination of affective, cognitive,
and behavioral factors and suicide-related outcomes in children and
young adolescents. Journal of Clinical Child and Adolescent Psychol-
ogy, 31, 48–58.
Nock, M. K., & Prinstein, M. J. (2004). A functional approach to the
assessment of self-mutilative behavior. Journal of Consulting and Clin-
ical Psychology, 72, 885–890.
Noshpitz, J. D. (1994). Self-destructiveness in adolescence. American
Journal of Psychotherapy, 48, 330–346.
Offer, D., & Barglow, P. (1960). Adolescent and young adult self-
mutilation incidents in a general psychiatric hospital. Archives of Gen-
eral Psychiatry, 3, 194–204.
Penn, J. V., Esposito, C. L., Schaeffer, L. E., Fritz, G. K., & Spirito, A.
(2003). Suicide attempts and self-mutilative behavior in a juvenile
correctional facility. Journal of the American Academy of Child and
Adolescent Psychiatry, 42, 762–769.
Prinstein, M. J., Nock, M. K., Spirito, A., & Grapentine, W. L. (2001).
Multimethod assessment of suicidality in adolescent psychiatric inpa-
tients: Preliminary results. Journal of the American Academy of Child
and Adolescent Psychiatry, 40, 1053–1061.
Ross, S., & Heath, N. (2002). A study of the frequency of self-mutilation
in a community sample of adolescents. Journal of Youth and Adoles-
cence, 31, 67–77.
Russell, D., Peplau, L. A., & Cutrona, C. E. (1980). The revised UCLA
Loneliness Scale: Concurrent and discriminant validity evidence. Jour-
nal of Personality and Social Psychology, 39, 472–480.
Schwab-Stone, M. E., Shaffer, D., Dulcan, M. K., Jensen, P. S., Fisher, P.,
Bird, H. R., et al. (1996). Criterion validity of the NIMH. Diagnostic
Interview Schedule for Children Version 2.3 (DISC–2.3). Journal of the
American Academy of Child and Adolescent Psychiatry, 35, 878–888.
Shaffer, D., Fisher, P., Dulcan, M. K., Davies, M., Piacentini, J., Schwab-
Stone, M. E., et al. (1996). The NIMH Diagnostic Interview Schedule
for Children Version 2.3 (DISC–2.3): Description, acceptability, preva-
lence rates, and performance in the MECA Study. Methods for the
epidemiology of Child and Adolescent Mental Disorders Study. Journal
of the American Academy of Child and Adolescent Psychiatry, 35,
Stanley, B., Gameroff, M. J., Michalsen, V., & Mann, J. J. (2001). Are
suicide attempters who self-mutilate a unique population? American
Journal of Psychiatry, 158, 427–432.
Suyemoto, K. L. (1998). The functions of self-mutilation. Clinical Psy-
chology Review, 18, 531–554.
van der Kolk, B. A., Perry, J. C., & Herman, J. L. (1991). Childhood
origins of self-destructive behavior. American Journal of Psychiatry,
Walker, R. L., Joiner, T. E., Jr., & Rudd, M. D. (2001). The course of
post-crisis suicidal symptoms: How and for whom is suicide “cathartic”?
Suicide and Life Threatening Behavior, 31, 144–152.
Walsh, B. W., & Rosen, P. M. (1988). Self-mutilation: Theory, research,
and treatment. New York: Guilford Press.
Zlotnick, C., Shea, M. T., Pearlstein, T., Simpson, E., Costello, E., &
Begin, A. (1996). The relationship between dissociative symptoms,
alexithymia, impulsivity, sexual abuse, and self-mutilation. Comprehen-
sive Psychiatry, 37, 12–16.
Received January 8, 2004
Revision received May 6, 2004
Accepted July 14, 2004 ?
NOCK AND PRINSTEIN