Prevalence of and Risk Factors for Suicide Attempts Versus Suicide
Gestures: Analysis of the National Comorbidity Survey
Matthew K. Nock
Ronald C. Kessler
Harvard Medical School
Definitions and classification schemes for suicide attempts vary widely among studies, introducing
conceptual, methodological, and clinical problems. We tested the importance of the intent to die criterion
by comparing self-injurers with intent to die, suicide attempters, and those who self-injured not to die but
to communicate with others, suicide gesturers, using data from the National Comorbidity Survey (n ?
5,877). Suicide attempters (prevalence ? 2.7%) differed from suicide gesturers (prevalence ? 1.9%) and
were characterized by male gender, fewer years of education, residence in the southern and western
United States; psychiatric diagnoses including depressive, impulsive, and aggressive symptoms; comor-
bidity; and history of multiple physical and sexual assaults. It is possible and useful to distinguish
between self-injurers on the basis of intent to die.
Keywords: suicide attempt, suicide gesture, intent to die, self-harm, self-injury
Suicide is among the leading causes of death around the world
(DeLeo, Bertolote, & Lester, 2002). It has been estimated that
approximately 4.6% of individuals in the United States have made
at least one suicide attempt in their lifetime (Kessler, Borges, &
Walters, 1999), and a prior suicide attempt is among the best
predictors of eventual death by suicide (Blumenthal, Bell, Neu-
mann, Schuttler, & Vogel, 1989; Goldstein, Black, Nasrallah, &
Winokur, 1991). Nevertheless, progress in the study of suicide has
been hindered by several key methodological problems. Perhaps
most important, there has been a lack of clarity and consistency in
the terms used to define suicide attempts. The different approaches
used for defining and classifying suicide attempts can be separated
into three different perspectives, differing on the treatment of the
presence of intent to die in the self-injurer. Many researchers and
clinicians use liberal criteria for defining suicide attempts, includ-
ing all self-injurious behavior, ignoring the issue of whether or not
there is intent to die (e.g., Lewinsohn, Rohde, & Seeley, 1996;
Seidlitz, Conwell, Duberstein, Cox, & Denning, 2001). Others
make a firm distinction between those with and without intent to
die and contend that the presence of lethal intent conveys signif-
icantly increased risk of death and should be explicitly assessed
and used in the classification of cases. Yet a third perspective is
that it is not useful or even possible to distinguish between self-
injurers with and without intent to die. From this perspective,
general terms such as parasuicide (Kreitman, 1977; Linehan,
1993; Platt et al., 1992) or deliberate self-harm (Hawton, Fagg,
Simkin, Bale, & Bond, 1997, 2000) are used to refer to all nonfatal
Emerging evidence suggests that it is indeed useful to distin-
guish between those with intent to die and those without such
intent. For instance, those with intent to die have been shown to
engage in more lethal self-injury (Beck, Beck, & Kovacs, 1975;
Brown, Henriques, Sosdjan, & Beck, 2004) and are more likely to
subsequently die by suicide (Harriss, Hawton, & Zahl, 2005;
Hjelmeland, 1996; Lonnqvist & Ostamo, 1991; Ostamo, Lon-
nqvist, Heinonen, Leppavuori et al., 1991). Consistent with this
perspective, an expert panel recently defined suicide attempts as
“potentially self-injurious behavior with a nonfatal outcome, for
which there is evidence (either implicit or explicit) that the person
intended at some (nonzero) level to kill himself/herself”
(O’Carroll, Berman, Maris, & Moscicki, 1996, p. 247). In contrast,
self-injury in which there is no intent to die, but instead an intent
to give the appearance of a suicide attempt in order to communi-
cate with others is commonly referred to as a suicide gesture.1
Despite these recommendations and supportive data, there contin-
ues to be definitional confusion and debate about the role of intent
in the definition of suicide attempts, and intent to die is rarely
explicitly assessed in studies of suicide attempts. This issue re-
quires resolution, as ignoring the intent of self-injury can lead to an
overestimation of the prevalence of suicide attempts and can
hinder the identification of risk factors specific to suicide attempts.
1O’Carroll et al. (1996) suggested the term instrumental suicide-related
behaviors for all self-injury in which there is no intent to die; however, we
use the term suicide gesture here given its greater specificity and more
common use in scientific and clinical communications.
Matthew K. Nock, Department of Psychology, Harvard University;
Ronald C. Kessler, Department of Health Care Policy, Harvard Medical
The National Comorbidity Survey (NCS) was supported by National
Institute of Mental Health Grants R01 MH46376, R01 MH49098, and R01
MH52861, with supplemental support from the National Institute of Drug
Abuse (through a supplement to MH46376) and the W. T. Grant Founda-
tion Grant 90135190. A complete list of NCS publications and the full text
of all NCS instruments can be found at http://www.hcp.med.harvard.edu/
ncs. Completion of the current study was supported in part by a grant from
the William A. Talley Fund of Harvard University.
Correspondence concerning this article should be addressed to Matthew
K. Nock, Department of Psychology, Harvard University, 33 Kirkland
Street, 1280, Cambridge, MA 02138. E-mail: firstname.lastname@example.org
Journal of Abnormal Psychology
2006, Vol. 115, No. 3, 616–623
Copyright 2006 by the American Psychological Association
0021-843X/06/$12.00 DOI: 10.1037/0021-843X.115.3.616
The current study was designed to extend previous work in this
area in several ways and to overcome some of the methodological
limitations of prior research. The first goal of this study was to
generate an estimate of the prevalence of lifetime suicide attempts
that explicitly considers intent to die, as well as an estimate of the
prevalence of suicide gestures, which has not been previously
reported in the literature. In addition, whereas most studies of
suicide attempts have used relatively small, selective samples,
limiting the generality of the findings of these studies, the current
study used data from the National Comorbidity Survey (NCS), a
nationally representative sample of individuals, which allows for
greater generality of the results than most previous studies of
suicide attempts. We expected the lifetime prevalence of suicide
attempts in this study to be substantially lower than the 4.6% figure
generated previously using a more liberal definition of suicide
attempt that did not explicitly require intent to die (Kessler et al.,
The second goal of the current study was to examine whether
those who report engaging in suicide attempt(s) with intent to die
(subsequently referred to as suicide attempts) differ significantly
from those without such intent, but with the intent of communi-
cating with others (subsequently referred to as suicide gestures).
Demonstrating differences between these two groups would serve
two important purposes. First, it would add to previous research
suggesting that self-injurers with intent to die differ from those
without such intent. Second, and perhaps more important, it would
result in the identification of risk factors specific to suicide at-
tempts (i.e., vs. suicide gestures). Virtually all studies examining
risk factors for suicide attempts have compared suicide attempters
to individuals with no history of self-injury. Therefore, it is unclear
if the risk factors identified in previous studies are associated with
self-injury in general or with suicide attempts in particular. Iden-
tifying risk factors specific to suicide attempts is an especially
important goal given the higher medical lethality and probability
of death associated with suicide attempts compared with suicide
gestures, as mentioned above.
Building on previous work in this area, we expected suicide
attempters to differ from those engaging in suicide gestures on
multiple domains, with the former group more closely resembling
those who die by suicide according to previous research in this
area. More specifically, we expected risk factors for suicide at-
tempts to be male gender, White race/ethnicity, fewer years of
education, and residence in the southern and western regions of the
United States, as each of these factors is associated with an
elevated rate of death by suicide (American Psychiatric Associa-
tion [APA], 2003; Centers for Disease Control [CDC], 2004;
DeLeo et al., 2002; Moscicki, 1999). In addition, we expected
suicide attempts to be uniquely predicted by the presence of
diagnoses most closely related to suicide death, such as affective
disorders, substance use disorders, conduct disorder, Cluster B
personality disorder, and psychiatric comorbidity (Brent et al.,
1988, 1993; Cavanagh, Carson, Sharpe, & Lawrie, 2003; Cheng,
Chen, Chen, & Jenkins, 2000; Shaffer et al., 1996). Finally, we
expected suicide attempts to be significantly and uniquely pre-
dicted by a history of physical and sexual abuse, factors that also
have been associated with elevated risk of suicide death in some
previous studies (Brent, Baugher, Bridge, Chen, & Chiappetta,
1999; Brent, Perper et al., 1994), but have not received as much
empirical consideration as demographic and diagnostic risk
The NCS was conducted among a nationally representative sample of
8,098 respondents aged 15 to 54 years interviewed between 1990 and 1992
(82.4% response rate). Informed consent was obtained from all respondents
and from the parents of all minors. Data were collected using a two-part
face-to-face structured interview completed in the home of each respon-
dent. Part 1 included the assessment of Diagnostic and Statistical Manual
of Mental Disorders (3rd ed. rev.; DSM–III–R; APA, 1987) disorders and
was administered to all respondents. Part 2 assessed risk factors and
consequences of the disorders evaluated in Part I, including all questions
about suicide attempts/gestures, and was administered to all respondents
who screened positive for any lifetime disorder in Part I as well as to a
probability subsample of other respondents. The current study includes the
5,877 respondents who participated in Part II. Data for these respondents
were weighted to correct for differential probabilities of selection into Part
II as well as for differential probabilities of within-household selection and
nonresponse. Comparisons of the demographic distribution of these 5,877
respondents with census data demonstrate that the sample is representative
of the U.S. population on a wide range of sociodemographic variables
(Kessler, Sonnega, Bromet, Hughes, & Nelson, 1995). Further details
about the design and methods used in the NCS are reported elsewhere
(Kessler, Little, & Groves, 1995; Kessler et al., 1994).
Assessment of suicide attempts/gestures.
asked questions about lifetime history of suicide attempts (“Have you ever
attempted suicide?”). Positive responses were followed with probes regard-
ing age of first attempt and the specific intent of the behavior. In cases in
which there was more than one such episode, the intent of the first and most
recent episodes was examined. Intent was assessed by asking respondents
which of the following three statements most accurately described their
attempt: (a) “I made a serious attempt to kill myself and it was only luck
that I did not succeed,” (b) “I tried to kill myself, but I knew the method
was not fool-proof,” and (c) “My attempt was a cry for help, I did not want
to die.” For the purposes of this report, we defined suicide attempters as
those with a lifetime history of ever endorsing Response a or b (n ? 156),
consistent with the definition described above requiring some (nonzero)
intent to die. Those who endorsed c (n ? 112), but never endorsed a or b,
were classified as having engaged in a suicide gesture.
Assessment of sociodemographic variables.
variables included in the analyses are sex, race/ethnicity, age, years of
education, religious affiliation, and current region of residence in the
United States. All sociodemographic characteristics of the NCS sample
closely match that of the U.S. population and are reported in full detail
elsewhere (Kessler, Sonnega et al., 1995).
Assessment of psychiatric diagnosis.
tained using a modified version of the Composite International Diagnostic
Interview (CIDI; Robins et al., 1988; World Health Organization [WHO],
1990b), a structured interview that generates diagnoses according to the
definitions and criteria of both the Diagnostic and Statistical Manual of
Mental Disorders (4th ed.; APA, 1994) and the ICD-10 (WHO, 1990a).
Field trials have supported the reliability and validity of all diagnoses
assessed by the CIDI (Kessler et al., 1998; Wittchen, 1994), with the
exception of mania and nonaffective psychosis (NAP) (Kendler, Gallagher,
Abelson, & Kessler, 1996; Kessler, Rubinow, Holmes, Abelson, & Zhao,
1997). Only the euphoric-grandiose subtype of mania, which is validly
assessed with the CIDI (Kessler, Davis, & Kendler, 1997), was included to
All Part II respondents were
Psychiatric diagnoses were ob-
INTENT TO DIE & SUICIDE
define mania in these analyses. Given that the CIDI overdiagnoses NAP
(Kessler et al., 1998), the CIDI was used as a first-stage screen for
psychosis and all respondents screening positive were reinterviewed in
order to confirm a full diagnosis of NAP (Kendler et al., 1996). The
presence and age of onset of each psychiatric diagnosis were used in the
Assessment of history of physical and sexual abuse.
history variables known to be associated with suicide were evaluated in
Part II of the NCS interview and were used in the current study. Respon-
dents were provided with a list of various traumatic events and asked to
indicate which events they had experienced, and at what age. Given past
research on risk factors for suicide, questions relevant for the current study
inquired about sexual assault (“You were raped [someone had sexual
intercourse with you when you did not want to by threatening you or using
some degree of force]”), (“You were sexually molested [someone touched
or felt your genitals when you did not want them to]”), physical assault
(“You were physically attacked or assaulted”), and childhood history of
physical abuse (“You were physically abused as a child”) and neglect
(“You were seriously neglected as a child”).
Each respondent was classified into one of three categories: no lifetime
history of self-injury, lifetime history of suicide gesture, or lifetime history
of suicide attempt. First, we calculated the prevalence of engaging in at
least one lifetime suicide attempt. Second, we calculated the prevalence of
each of the sociodemographic, diagnostic, and abuse history variables and
compared them across the suicide gesture and suicide attempt groups. We
do not provide an overall comparison of non-self-injurious with self-
injurious individuals in this study, as these data from the NCS have been
presented previously (Kessler et al., 1999). Group differences in the prev-
alence of potential risk factors were evaluated using simple cross tabula-
tions. We evaluated the effect of each of these variables in distinguishing
suicide attempters from suicide gesturers. These analyses were conducted
using logistic regression models. Results are reported in odds ratios (and
95% confidence intervals) obtained by exponentiating the regression co-
efficients from the logistic models. Because we were interested in identi-
fying risk factors for suicide attempts, we included in our analyses only
those demographic, diagnostic, and abuse events that occurred prior to the
self-injury in question. That is, for each respondent only those diagnoses
and events that occurred prior to that individual’s earliest identified self-
injury were included. This provided a more conservative test of the study
hypotheses and ensures the temporal precedence of all risk factors exam-
ined. As mentioned, data were weighted to correct for selection and
nonresponse. In evaluating the results of our analyses we used conventional
statistical significance testing, rather than design-based adjustments, given
that our goal was to evaluate specific hypotheses rather than to make
Prevalence of Suicide Attempts and Suicide Gestures
A total of 268 respondents (4.6% of the sample) reported a
lifetime history of “suicide attempt.” However, requiring intent to
die in the definition of suicide attempt reduced that number to 156
(2.7% of the sample). Subsequent analyses are based on compar-
isons between those with a lifetime history of a true suicide
attempt versus the 112 (1.9% of the sample) respondents with a
history of suicide gesture.
The majority of respondents who endorsed making either a
suicide attempt or suicide gesture reported only one such act in
their lifetime (69.8%), and the rate of repeated self-injury did not
differ significantly among suicide attempters (34.6%) and suicide
gesturers (24.1%).2In cases of repeated self-injury, the NCS
examined respondents’ intent during only the first and most recent
self-injurious act, precluding a fine-grained analysis of each epi-
sode. Nevertheless, there was consistency in the reported intent,
with 70.8% of repeated self-injurers reporting the same intent for
both episodes and only 29.2% reporting a change in intent.
Sociodemographic Risk Factors for Suicide Attempt
Sociodemographic risk factors for intent to die are presented in
Table 1. Among all self-injurers, specific risk factors for the
presence of suicide attempt were male gender, fewer years of
education, and residence in the southern and western United
States. Individuals without a specific religious affiliation were
twice as likely to engage in a suicide attempt, ?1
however, this risk was not statistically significant in the logistic
regression equation (p ? .060). All statistically significant socio-
demographic risk factors were statistically controlled in all subse-
2? 4.21, p ? .040;
Prior Psychiatric Disorders as Risk Factors for Suicide
The prevalence of each psychiatric diagnosis assessed is re-
ported for suicide gesturers and suicide attempters in Table 2.
Specific risk factors for suicide attempt included the prior presence
of psychiatric diagnoses associated with affective, impulsive, and
aggressive symptoms. More specifically, the presence of a major
depressive episode, drug dependence or abuse, conduct disorder,
and antisocial personality disorder, as well as the presence of
psychiatric comorbidity (?3 disorders), all significantly increased
the risk of suicide attempt. The presence of mania increased the
likelihood of suicide attempt ninefold (i.e., eight of the nine
self-injurers who met diagnostic criteria for mania reported a
suicide attempt); however, given the low prevalence of mania this
relation was not statistically significant, ?1
Simple phobia also emerged as a risk factor for suicide attempt.
Overall, several specific diagnoses and a greater accumulation of
diagnoses (i.e., psychiatric comorbidity) distinguished suicide at-
tempters from suicide gesturers.
2? 3.64, p ? .057.
Physical and Sexual Assault as Risk Factors for Suicide
The prevalence of physical and sexual abuse between the two
groups is presented in Table 3. Although the presence of rape or
sexual molestation did not differ between suicide gesturers and
attempters, the risk of suicide attempt was significantly increased
in the presence of multiple rapes and multiple sexual molestations
as well as with higher rates of physical assault. Physical abuse and
2The NCS also examined the total number of lifetime “suicide attempts”
reported regardless of intent to die. The results of each of the subsequent
regression analyses were virtually unchanged when this variable was
entered in the first step of each regression equation demonstrating the
results are not accounted for by the total number of such acts. Given this,
as well as the questionable validity and reliability of this variable due to
uncertainty about the intent to die in each episode, the results are reported
without this variable in the equations.
NOCK AND KESSLER
neglect during childhood were not associated with increased risk
of suicide attempt.
Individuals who engage in self-injury with intent to die differ
significantly from those without such intent. Risk factors for
membership in the former group closely resemble those for suicide
death. These results add to previous findings demonstrating that
those with intent to die from self-injury are more likely to sustain
more medically lethal injuries and are more likely to ultimately die
by suicide (Brown et al., 2004; Harriss et al., 2005). Taken
together, these findings highlight the importance of distinguishing
between self-injurers with and without intent to die. As such, it is
recommended that researchers and clinicians avoid using terms
and classification schemes that ignore or obscure these differences,
such as parasuicide and deliberate self-harm, and instead classify
self-injury according to the intent of such behavior.
Previous epidemiologic studies have reported lifetime preva-
lence of suicide attempts as 1.1% to 4.6% (Kessler et al., 1999;
Moscicki et al., 1988; Paykel, Myers, Lindenthal, & Tanner,
1974). These varying rates may be the result of different wording
or interpretation of assessment items. An advantage of the NCS
and a strength of the current study is the inclusion of follow-up
items regarding the specific intent of individuals’ self-injury. The
current study revealed that although 4.6% of respondents indicated
they made a “suicide attempt,” only 2.7% reported doing so with
intent to die, whereas 1.9% denied any such intent and reported
doing so to communicate with others. Given that almost half (42%)
of those who reported making a “suicide attempt” indicated they
had no intent to die, we strongly recommend that future studies
explicitly assess and require intent to die as a criterion for defining
suicide attempts. We used a general classification scheme for
intent to die based on presence or absence of this construct. Future
studies may benefit from the use of more fine-grained analyses of
intent to die, such as the examination of intent to die as a contin-
uous measure (e.g., Beck et al., 1975). Explicitly evaluating the
intent of self-injury will advance research in several ways. Assess-
ing intent facilitates understanding of the factors maintaining self-
injurious behavior (e.g., Nock & Prinstein, 2004, 2005). In addi-
tion, carefully defining key constructs, such as suicide attempts,
for participants and patients will reduce variation in responding
and will enhance interpretation of study results (see Linehan, 1997;
Meehan, Lamb, Saltzman, & O’Carroll, 1992).
Emerging evidence demonstrates that certain risk factors are
useful for predicting the presence of self-injurious thoughts and
behaviors in general, whereas other risk factors are useful in
distinguishing among specific types of self-injurious thoughts and
behaviors. For instance, in one recent study, individuals’ level of
Demographic Characteristics by Presence of Intent to Die
% OR (95% CI)
Less than high school
High school graduate
Some post-HS (13–15 yrs)
College graduate (16 yrs)
Region of U.S.A.
0.07 (0.2–2.0) 11.82**
variable) with suicide attempt status as the dependent variable. HS ? high school.
* p ? .05. ** p ? .01.
Odds ratios (OR) and confidence intervals (CI) were obtained from logistic regression equations (one per
INTENT TO DIE & SUICIDE
depressed mood distinguished currently suicidal (ideation or at-
tempts) from nonsuicidal individuals, but only the presence of past
suicide attempts and anhedonia distinguished between current sui-
cidal ideators and suicide attempters (Nock & Kazdin, 2002).
Identifying specific risk factors for different types of suicide-
related outcomes is necessary to improve the accuracy of the
prediction of such outcomes. Toward this end, the current findings
demonstrate that multiple factors distinguish between suicide ges-
turers and suicide attempters, including demographic, diagnostic,
and abuse history variables.
Consistent with previous reports, more women than men in this
study engaged in self-injury in general. However, men who en-
gaged in self-injury were more likely to make suicide attempts
than suicide gestures, whereas women were more likely to make
suicide gestures than suicide attempts. In other words, men were
more likely to report intent to die from their self-injury, whereas
women were more likely to report doing so as a means of com-
municating with others. This finding is in line with earlier sugges-
tions that men are more likely than women to die by suicide
because of a higher likelihood of lethal intent (Denning, Conwell,
King, & Cox, 2000), and with findings that the rate of suicide
attempts for men and women becomes approximately equal if one
considers only medically serious suicide attempts (Beautrais,
2001). Two important caveats deserve serious consideration. First,
it is possible that these results reflect a reporting bias and there was
no actual difference in true intent at the time of self-injury.
However, the fact that men are more likely than women to use
lethal methods (e.g., firearms) during suicide attempts (CDC,
2004) suggests this finding may reflect a true gender difference.
Second, even though intent to die is associated with medical
lethality and subsequent death by suicide (e.g., Harriss et al.,
2005), self-injury in the absence of intent to die is a very dangerous
behavior requiring serious research and clinical attention. The
absence of intent to die does not in itself protect against death from
We also found that psychiatric diagnoses associated with de-
pressive (Major Depressive Episode and Mania), impulsive (Drug
Abuse and Dependence), and aggressive (Conduct Disorder and
Antisocial Personality Disorder) behaviors increased the risk of
suicide attempt. These diagnoses, along with psychiatric comor-
bidity, are among the most consistently supported risk factors for
suicide death (Brent, Johnson et al., 1994; Brent et al., 1993;
Cavanagh et al., 2003; Cheng et al., 2000; DeLeo et al., 2002;
Foster, Gillespie, McClelland, & Patterson, 1999; Henriksson et
al., 1993; Isometsa et al., 1996; Shaffer et al., 1996; Shafii,
Steltz-Lenarsky, Derrick, Beckner, & Whittinghill, 1988). Thus,
although the presence of any psychiatric disorder increases the risk
of self-injurious thoughts and behaviors generally (Kessler et al.,
1999), researchers and clinicians should be especially careful to
monitor the likelihood of suicide attempts in the presence of these
specific diagnoses. Simple phobia was also a risk factor for suicide
attempt in the current study. Prior studies have reported a relation
between acute anxiety/agitation and suicide attempts (e.g., Faw-
cett, 2001). However, previous support for the relation between
simple phobia and suicide attempts is lacking and the current
finding may be a function of the relation between simple phobia
and comorbid disorders that are themselves independently associ-
ated with suicide attempts (e.g., Vickers & McNally, 2004).
Psychiatric diagnoses may be best conceptualized as proximal
risk factors for suicide attempts. A growing body of literature
highlights the relation between more distal risk factors, such as
physical and sexual abuse, and the occurrence of self-mutilation
(Gratz, 2003; van der Kolk, Perry, & Herman, 1991), suicide
attempts (Brodsky, Malone, Ellis, Dulit, & Mann, 1997; Brodsky
et al., 2001; Dube et al., 2001), and suicide death (Brent et al.,
1999). What has been less clear is whether and how a history of
traumatic events is differentially related to these different forms of
Lifetime Prevalence of Physical and Sexual Abuse and Intent
% OR (95% CI)
History of “rape”
History of “sexual molestation”
History of physical assault
Physical abuse as child
Neglect as child
logistic regression equations (one per row) with suicide attempt status as
the dependent variable, controlling for demographic variables from Ta-
* p ? .05, by two-sided test.
Odds ratios (OR) and confidence intervals (CI) were obtained from
Lifetime Prevalence of DSM Diagnoses and Presence of Intent
% OR (95% CI)
Major depressive episode
Antisocial personality disorder
Posttraumatic stress disorder
Generalized anxiety disorder
Any 1 disorder
Any 2 disorders
logistic regression equations (one per row) with suicide attempt status as
the dependent variable, controlling for significant demographic variables
from Table 1. DSM ? Diagnostic and Statistical Manual of Mental
* p ? .05, by two-sided test.** p ? .01.
Odds ratios (OR) and confidence intervals (CI) were obtained from
NOCK AND KESSLER
self-injury. Our findings again add specificity to previous reports.
It is interesting that an overall history of rape, sexual molestation,
and child abuse or neglect was unrelated to the presence of intent
to die; however, intent to die was uniquely associated with multiple
incidents of rape and sexual molestation, as well as with a history
of physical assault. These data suggest a dose–response relation
between sexual trauma and subsequent self-injury, with more
frequent trauma leading to suicide attempt. However, respondents
were not asked the specific number of traumatic incidents that
occurred, so it is unclear whether there is a linear dose–response
relation or if two sexually traumatic incidents are sufficient to
convey an increased risk of suicide attempt. Nevertheless, the
results indicate that repeated traumatic events carry an increased
risk of suicide attempt. The specificity of our results for sexual and
physical abuse, but not neglect, is consistent with previous find-
ings in this area (e.g., Green, 1978).
These findings should be viewed in the context of the limitations
of this study. First, the NCS is a cross-sectional survey relying on
retrospective self-report; therefore, the responses may have been
affected by individual biases and inaccuracies. For instance, re-
spondents may have forgotten events, made misattributions about
the reasons for their behaviors, or may have erred regarding the
timing of the events assessed. It is also possible that current or
more recent events or mood states may have biased individuals’
recollection of past events (see Schacter, 1999, for a review). Such
biases are unlikely to have affected the sociodemographic factors,
but may have influenced recall of the diagnostic and abuse history
variables. However, if such biases were present, they were limited
in scope, as we found specific effects for hypothesized diagnostic
and abuse variables. Second, many of the constructs were assessed
using a single-item rather than multiple-item scales, which may
limit the reliability of responses. In addition, although the abuse
history items provided specific definitions of each construct as-
sessed (e.g., “rape,” “molestation”) it is possible that individual or
idiosyncratic interpretations of these items introduced bias or error
into responses. Problems associated with recall and single-item
measurement are also likely to have influenced our assessment of
suicide attempts and gestures. It is interesting that suicide attempts
were more prevalent than suicide gestures, and this finding high-
lights the scope and the seriousness of the problem under investi-
gation. However, these estimates may have been affected by the
inclusion of only individuals who responded affirmatively to the
first item about making a “suicide attempt.” Not captured by these
assessment methods are individuals who would not label their
self-injurious behavior a “suicide attempt,” but may instead con-
sider it a suicide gesture or other form of nonsuicidal self-injury
(e.g., self-mutilation). Inclusion of such cases would have in-
creased the estimated prevalence of suicide gestures generated in
this study. As such, we recommend that future studies include
assessment of self-injury with and without intent to die among all
A third limitation of this study is the presence of measurement
error associated with current methods for assessing intent to die.
There is a growing literature demonstrating limitations in individ-
uals’ ability to accurately report their intentions for engaging in
different behaviors (Bargh, Gollwitzer, Lee-Chai, Barndollar, &
Trotschel, 2001; Wegner, 2004), and such limitations undoubtedly
apply to the assessment of intent to die. Although previous studies
and the current report demonstrate the usefulness of self-report of
intent to die, the development of more objective and hopefully
more accurate methods for evaluating this construct represent an
important goal for future work in this area.
Finally, the NCS did not collect data on individuals who died by
suicide; therefore, we were unable to directly compare the char-
acteristics and risk factors for suicide attempts with those of
suicide (see Beautrais, 2001), and instead relied on findings of
those who died by suicide reported in other studies. Nevertheless,
the current study provides an advance over previous studies by
delineating differences between suicide gesturers and attempters
and highlighting risk factors specific to suicide attempts.
In conclusion, we demonstrated that those who report engaging
in self-injury with intent to die differ in significant ways from
self-injurers without such intent. These results underline the im-
portance of using intent to die to define and classify self-injurers
and provide key information about the risk factors for engaging in
such behavior. Researchers and clinicians are encouraged to attend
to the importance of intent in making methodological and clinical
decisions surrounding self-injury, and to use clear and consistent
definitions for constructs related to self-injury. Although perceived
intent of one’s behavior is a complex and dynamic construct,
assessments of intent have proven useful in predicting severity of
self-injury and death in previous studies, and in revealing hypoth-
esized differences in risk factors in the current study. These find-
ings suggest that classifying individuals on the basis of the intent
of their self-injury is a useful scientific and clinical endeavor.
Several lines of future work follow directly from our findings,
including the use of longitudinal data to clarify temporal relations
between predictors and outcomes, the evaluation of more specific
risk factors for the various types of self-injury, and the evaluation
of models for the development and maintenance of self-injury that
provide a more thorough integration of the various risk factors
identified in this and previous studies.
American Psychiatric Association. (1987). Diagnostic and statistical man-
ual of mental disorders (3rd ed., rev.). Washington, DC: Author.
American Psychiatric Association. (1994). Diagnostic and statistical man-
ual of mental disorders (4th ed.). Washington, DC: Author.
American Psychiatric Association. (2003). Practice guideline for the as-
sessment and treatment of patients with suicidal behaviors. American
Journal of Psychiatry, 160(Suppl. 11), 1–60.
Bargh, J. A., Gollwitzer, P. M., Lee-Chai, A., Barndollar, K., & Trotschel,
R. (2001). The automated will: Nonconscious activation and pursuit of
behavioral goals. Journal of Personality and Social Psychology, 81,
Beautrais, A. L. (2001). Suicides and serious suicide attempts: Two pop-
ulations or one? Psychological Medicine, 31, 837–845.
Beck, A. T., Beck, R., & Kovacs, M. (1975). Classification of suicidal
behaviors: I. Quantifying intent and medical lethality. American Journal
of Psychiatry, 132, 285–287.
Blumenthal, S., Bell, V., Neumann, N. U., Schuttler, R., & Vogel, R.
(1989). Mortality and rate of suicide of first admission psychiatric
patients: A 5-year follow-up of a prospective longitudinal study. Psy-
chopathology, 22, 50–56.
Brent, D. A., Baugher, M., Bridge, J., Chen, T., & Chiappetta, L. (1999).
Age- and sex-related risk factors for adolescent suicide. Journal of the
American Academy of Child and Adolescent Psychiatry, 38, 1497–1505.
Brent, D. A., Johnson, B. A., Perper, J., Connolly, J., Bridge, J., Bartle, S.,
et al. (1994). Personality disorder, personality traits, impulsive violence,
INTENT TO DIE & SUICIDE
and completed suicide in adolescents. Journal of the American Academy
of Child and Adolescent Psychiatry, 33, 1080–1086.
Brent, D. A., Perper, J. A., Goldstein, C. E., Kolko, D. J., Allan, M. J.,
Allman, C. J., et al. (1988). Risk factors for adolescent suicide. A
comparison of adolescent suicide victims with suicidal inpatients. Ar-
chives of General Psychiatry, 45, 581–588.
Brent, D. A., Perper, J. A., Moritz, G., Allman, C., Friend, A., Roth, C., et
al. (1993). Psychiatric risk factors for adolescent suicide: A case-control
study. Journal of the American Academy of Child and Adolescent
Psychiatry, 32, 521–529.
Brent, D. A., Perper, J. A., Moritz, G., Liotus, L., Schweers, J., Balach, L.,
et al. (1994). Familial risk factors for adolescent suicide: A case-control
study. Acta Psychiatrica Scandanavica, 89, 52–58.
Brodsky, B. S., Malone, K. M., Ellis, S. P., Dulit, R. A., & Mann, J. J.
(1997). Characteristics of borderline personality disorder associated with
suicidal behavior. American Journal of Psychiatry, 154, 1715–1719.
Brodsky, B. S., Oquendo, M., Ellis, S. P., Haas, G. L., Malone, K. M., &
Mann, J. J. (2001). The relationship of childhood abuse to impulsivity
and suicidal behavior in adults with major depression. American Journal
of Psychiatry, 158, 1871–1877.
Brown, G. K., Henriques, G. R., Sosdjan, D., & Beck, A. T. (2004). Suicide
intent and accurate expectations of lethality: Predictors of medical
lethality of suicide attempts. Journal of Consulting and Clinical Psy-
chology, 72, 1170–1174.
Cavanagh, J. T., Carson, A. J., Sharpe, M., & Lawrie, S. M. (2003).
Psychological autopsy studies of suicide: A systematic review. Psycho-
logical Medicine, 33, 395–405.
Centers for Disease Control. (2004). Web-based Injury Statistics Query
and Reporting System (WISQARS). Retrieved January, 20, 2005.
Cheng, A. T., Chen, T. H., Chen, C. C., & Jenkins, R. (2000). Psychosocial
and psychiatric risk factors for suicide. Case-control psychological au-
topsy study. British Journal of Psychiatry, 177, 360–365.
DeLeo, D., Bertolote, J., & Lester, D. (2002). Self-directed violence. In
E. G. Krug, L. L. Dahlberg, J. A. Mercy, A. B. Zwi, & R. Lozano (Eds.),
World report on violence and health (pp. 183–240). Geneva, Switzer-
land: World Health Organization.
Denning, D. G., Conwell, Y., King, D., & Cox, C. (2000). Method choice,
intent, and gender in completed suicide. Suicide and Life Threatening
Behavior, 30, 282–288.
Dube, S. R., Anda, R. F., Felitti, V. J., Chapman, D. P., Williamson, D. F.,
& Giles, W. H. (2001). Childhood abuse, household dysfunction, and the
risk of attempted suicide throughout the life span: Findings from the
Adverse Childhood Experiences Study. Journal of the American Medi-
cal Association, 286, 3089–3096.
Fawcett, J. (2001). Treating impulsivity and anxiety in the suicidal patient.
Annals of the New York Academy of Sciences, 932, 94–105.
Foster, T., Gillespie, K., McClelland, R., & Patterson, C. (1999). Risk
factors for suicide independent of DSM–III–R Axis I disorder. Case-
control psychological autopsy study in Northern Ireland. British Journal
of Psychiatry, 175, 175–179.
Goldstein, R. B., Black, D. W., Nasrallah, A., & Winokur, G. (1991). The
prediction of suicide. Sensitivity, specificity, and predictive value of a
multivariate model applied to suicide among 1906 patients with affective
disorders. Archives of General Psychiatry, 48, 418–422.
Gratz, K. L. (2003). Risk factors for and functions of deliberate self-harm:
An empirical and conceptual review. Clinical Psychology: Science and
Practice, 10, 192–205.
Green, A. H. (1978). Self-destructive behavior in battered children. Amer-
ican Journal of Psychiatry, 135, 579–582.
Harriss, L., Hawton, K., & Zahl, D. (2005). Value of measuring suicidal
intent in the assessment of people attending hospital following self-
poisoning or self-injury. British Journal of Psychiatry, 186, 60–66.
Hawton, K., Fagg, J., Simkin, S., Bale, E., & Bond, A. (1997). Trends in
deliberate self-harm in Oxford, 1985–1995. Implications for clinical
services and the prevention of suicide. British Journal of Psychiatry,
Hawton, K., Fagg, J., Simkin, S., Bale, E., & Bond, A. (2000). Deliberate
self-harm in adolescents in Oxford, 1985–1995. Journal of Adolescence,
Henriksson, M. M., Aro, H. M., Marttunen, M. J., Heikkinen, M. E.,
Isometsa, E. T., Kuoppasalmi, K. I., et al. (1993). Mental disorders and
comorbidity in suicide. American Journal of Psychiatry, 150, 935–940.
Hjelmeland, H. (1996). Verbally expressed intentions of parasuicide: II.
Prediction of fatal and nonfatal repetition. Crisis, 17, 10–14.
Isometsa, E. T., Henriksson, M. M., Heikkinen, M. E., Aro, H. M.,
Marttunen, M. J., Kuoppasalmi, K. I., et al. (1996). Suicide among
subjects with personality disorders. American Journal of Psychiatry,
Kendler, K. S., Gallagher, T. J., Abelson, J. M., & Kessler, R. C. (1996).
Lifetime prevalence, demographic risk factors, and diagnostic validity of
nonaffective psychosis as assessed in a US community sample: The
National Comorbidity Survey. Archives of General Psychiatry, 53,
Kessler, R. C., Borges, G., & Walters, E. E. (1999). Prevalence of and risk
factors for lifetime suicide attempts in the National Comorbidity Survey.
Archives of General Psychiatry, 56, 617–626.
Kessler, R. C., Davis, C. G., & Kendler, K. S. (1997). Childhood adversity
and adult psychiatric disorder in the US National Comorbidity Survey.
Psychological Medicine, 27, 1101–1119.
Kessler, R. C., Little, R. J., & Groves, R. M. (1995). Advances in strategies
for minimizing and adjusting for survey nonresponse. Epidemiologic
Reviews, 17, 192–204.
Kessler, R. C., McGonagle, K. A., Zhao, S., Nelson, C. B., Hughes, M.,
Eshleman, S., et al. (1994). Lifetime and 12-month prevalence of DSM–
III–R psychiatric disorders in the United States: Results from the Na-
tional Comorbidity Survey. Archives of General Psychiatry, 51, 8–19.
Kessler, R. C., Rubinow, D. R., Holmes, C., Abelson, J. M., & Zhao, S.
(1997). The epidemiology of DSM–III–R bipolar I disorder in a general
population survey. Psychological Medicine, 27, 1079–1089.
Kessler, R. C., Sonnega, A., Bromet, E., Hughes, M., & Nelson, C. B.
(1995). Posttraumatic stress disorder in the National Comorbidity Sur-
vey. Archives of General Psychiatry, 52, 1048–1060.
Kessler, R. C., Wittchen, H. U., Abelson, J. M., McGonagle, K. A.,
Schwartz, N., Kendler, K. S., et al. (1998). Methodological studies of the
Composite International Diagnostic Interview (CIDI) in the U.S. Na-
tional Comorbidity Survey. International Journal of Methods in Psychi-
atric Research, 7, 33–55.
Kreitman, N. (1977). Parasuicide. London: Wiley.
Lewinsohn, P. M., Rohde, P., & Seeley, J. R. (1996). Adolescent suicide
ideation and attempts: Prevalence, risk factors, and clinical implications.
Clinical Psychology: Science and Practice, 3, 25–46.
Linehan, M. M. (1993). Cognitive–behavioral treatment of borderline
personality disorder. New York: Guilford Press.
Linehan, M. M. (1997). Behavioral treatments of suicidal behaviors: Def-
initional obfuscation and treatment outcomes. Annals of the New York
Academy of Sciences, 836, 302–328.
Lonnqvist, J., & Ostamo, A. (1991). Suicide following the first suicide
attempt: A five-year follow-up using a survival analysis. Psychiatria
Fennica, 22, 171–179.
Meehan, P. J., Lamb, J. A., Saltzman, L. E., & O’Carroll, P. W. (1992).
Attempted suicide among young adults: Progress toward a meaningful
estimate of prevalence. American Journal of Psychiatry, 149, 41–44.
Moscicki, E. K. (1999). Epidemiology of suicide. In D. G. Jacobs (Ed.),
The Harvard Medical School guide to suicide assessment and interven-
tion (pp. 40–51). San Francisco: Jossey-Bass.
Moscicki, E. K., O’Carroll, P., Rae, D. S., Locke, B. Z., Roy, A., & Regier,
D. A. (1988). Suicide attempts in the Epidemiologic Catchment Area
Study. Yale Journal of Biological Medicine, 61, 259–268.
NOCK AND KESSLER
Nock, M. K., & Kazdin, A. E. (2002). Examination of affective, cognitive, Download full-text
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.
Nock, M. K., & Prinstein, M. J. (2005). Clinical features and behavioral
functions of adolescent self-mutilation. Journal of Abnormal Psychol-
ogy, 114, 140–146.
O’Carroll, P. W., Berman, A. L., Maris, R., & Moscicki, E. (1996). Beyond
the tower of Babel: A nomenclature for suicidology. Suicide & Life-
Threatening Behavior, 26, 237–252.
Ostamo, A., Lonnqvist, J., Heinonen, S., Leppavuori, A., et al. (1991).
Epidemiology of parasuicides in Finland. Psychiatria Fennica, 22, 181–
Paykel, E. S., Myers, J. K., Lindenthal, J. J., & Tanner, J. (1974). Suicidal
feelings in the general population: A prevalence study. British Journal of
Psychiatry, 124, 460–469.
Platt, S., Bille-Brahe, U., Kerkhof, A., Schmidtke, A., Bjerke, T., Crepet,
P., et al. (1992). Parasuicide in Europe: The WHO/EURO multicentre
study on parasuicide: I. Introduction and preliminary analysis for 1989.
Acta Psychiatrica Scandanavica, 85, 97–104.
Robins, L. N., Wing, J., Wittchen, H. U., Helzer, J. E., Babor, T. F., Burke,
J., et al. (1988). The Composite International Diagnostic Interview: An
epidemiologic instrument suitable for use in conjunction with different
diagnostic systems and in different cultures. Archives of General Psy-
chiatry, 45, 1069–1077.
Schacter, D. L. (1999). The seven sins of memory: Insights from psychol-
ogy and cognitive neuroscience. American Psychologist, 54, 182–203.
Seidlitz, L., Conwell, Y., Duberstein, P., Cox, C., & Denning, D. (2001).
Emotion traits in older suicide attempters and non-attempters. Journal of
Affective Disorders, 66, 123–131.
Shaffer, D., Gould, M. S., Fisher, P., Trautman, P., Moreau, D., Kleinman,
M., et al. (1996). Psychiatric diagnosis in child and adolescent suicide.
Archives of General Psychiatry, 53, 339–348.
Shafii, M., Steltz-Lenarsky, J., Derrick, A. M., Beckner, C., & Whitting-
hill, J. R. (1988). Comorbidity of mental disorders in the post-mortem
diagnosis of completed suicide in children and adolescents. Journal of
Affective Disorders, 15, 227–233.
van der Kolk, B. A., Perry, J. C., & Herman, J. L. (1991). Childhood
origins of self-destructive behavior. American Journal of Psychiatry,
Vickers, K., & McNally, R. J. (2004). Panic disorder and suicide attempt
in the National Comorbidity Survey. Journal of Abnormal Psychology,
Wegner, D. M. (2004). Precis of the illusion of conscious will. Behavioral
and Brain Sciences, 27, 649–692.
World Health Organization. (1990a). International Classification of
Diseases–Classification of mental and behavioral disorders: Diagnostic
criteria for research. Geneva, Switzerland: Author.
World Health Organization. (1990b). Composite International Diagnostic
Interview: Version 1.0. Geneva, Switzerland: Author.
Wittchen, H. U. (1994). Reliability and validity studies of the WHO–
Composite International Diagnostic Interview (CIDI): A critical review.
Journal of Psychiatric Research, 28, 57–84.
Received September 13, 2004
Revision received July 28, 2005
Accepted August 1, 2005 ?
INTENT TO DIE & SUICIDE