Suicide Ideation and Attempts among Inhalant Users: Results from the National Epidemiologic Survey on Alcohol and Related Conditions

Article (PDF Available)inSuicide and Life-Threatening Behavior 40(3):276-86 · June 2010with30 Reads
DOI: 10.1521/suli.2010.40.3.276 · Source: PubMed
Few studies have examined associations of inhalant use and inhalant use disorders (IUDs) to suicide ideation and attempts. We investigated these relationships in the largest comorbidity survey conducted in the United States. Suicidal ideation was significantly more prevalent among inhalant users than nonusers and severity of inhalant use problems was positively related to suicidal ideation. Among persons with IUDs, 67.4% had thought about committing suicide and 20.2% had attempted suicide. Multivariate logistic regression analyses indicated that respondents with IUDs reported significantly higher levels of suicide ideation than inhalant nonusers. Inhalant use is associated with significantly increased risk for suicide ideation, especially among women and persons with DSM-IV IUDs.
276 Suicide and Life-Threatening Behavior 40(3) June 2010
2010 The American Association of Suicidology
Suicide Ideation and Attempts among
Inhalant Users: Results from the
National Epidemiologic Survey
on Alcohol and Related Conditions
Matthew O. Howard, PhD, Brian E. Perron, PhD, Paul Sacco, PhD,
Mark Ilgen, PhD, Michael G. Vaughn, PhD, Eric Garland, MSW,
and Stacey Freedentahl, PhD
Few studies have examined associations of inhalant use and inhalant use
disorders (IUDs) to suicide ideation and attempts. We investigated these relation-
ships in the largest comorbidity survey conducted in the United States. Suicidal
ideation was significantly more prevalent among inhalant users than nonusers and
severity of inhalant use problems was positively related to suicidal ideation.
Among persons with IUDs, 67.4% had thought about committing suicide and
20.2% had attempted suicide. Multivariate logistic regression analyses indicated
that respondents with IUDs reported significantly higher levels of suicide ideation
than inhalant nonusers. Inhalant use is associated with significantly increased risk
for suicide ideation, especially among women and persons with DSM-IV IUDs.
Psychiatric disorders are associated with in- chiatric disorders, substance use disorders are
among the strongest known risk factors forcreased risk for suicidal behavior. A psychiat-
ric diagnosis is present in 90% or more of completed suicide (Wilcox, Conner, & Caine,
2004). According to a review of cohort stud-deaths from suicide in western countries in-
cluding the United States (Bertolote, Fleisch- ies, individuals with alcohol dependence are
at 9.8 (CI: 9.0 to 10.7) times greater risk tomann, De Leo, & Wasserman, 2004). Of psy-
die by suicide compared to the general popu-
lation, and persons with other drug use disor-
Matthew Howard and Eric Garland are
ders are at 13.7 (CI: 10.3 to 18.0) times
with the University of North Carolina at Chapel
greater risk (Wilcox et al., 2004). Substance
Hill; Brian Perron and Mark Ilgen are with the
use disorders are also highly predictive of sui-
University of Michigan at Ann Arbor; Paul Sacco
cidal thoughts, plans, and nonfatal attempts
is with Washington University; Michael Vaughn
is with Saint Louis University; and Stacey
(Borges, Walters, & Kessler, 2000; Wilcox &
Freedenthal is with the University of Denver.
Anthony, 2004). Substance use may directly
This study was supported by grants
induce lower serotonergic central nervous
DA021405 (Natural History, Comorbid Mental
system activity and produce dysfunction in
Disorders, and Consequences of Inhalant Abuse,
the ventral prefrontal cortex, an area respon-
M.O. Howard, PI), and DA15929 (Neuropsychi-
atric Impairment in Adolescent Inhalant Abusers,
sible for inhibition of aggressive impulses (in-
M.O. Howard, PI) from the National Institute on
cluding suicide attempts). Substance use may
Drug Abuse.
also contribute indirectly to suicide risk via
Address corresponde nce to Dr. Matthew O.
precipitation of psychosocial crises and con-
Howard, 325 Pittsboro, CB #3550, The University
sequent feelings of dysphoria and hopeless-
of North Carolina at Chapel Hill, Chapel Hill, NC
27599; E-mail: mohoward@ emai l.u nc.e du
ness (Mann, 2003).
Howard et al. 277
Relatively few studies have examined (2004) identified significantly elevated rates
of lifetime suicide attempt among adolescentsuicide risk in relation to specific psychoac-
tive substances of abuse, but available reports inhalant users with and without DSM-IV
IUDs (32% vs. 27%, respectively) comparedsuggest that inhalants may be unique in
terms of the frequency and degree to which to inhalant nonusers (15%), but the differ-
ence between inhalant users with and withouttheir use is associated with enhanced risk for
suicide ideation and attemptsespecially among IUDs in lifetime suicide attempt prevalence
was not statistically significant.girls and women (Luncheon, Bae, Gonzalez,
Lurie, & Singh, 2008; Wilcox & Anthony, There are several possible explanations
for reported findings identifying significant2004). Borges et al. (2000) assessed the rela-
tionship of substance use, abuse without de- associations between inhalant use, IUDs and
suicide ideation and attempts. Inhalation ofpendence, and dependence to risk for subse-
quent first suicide attempt across ten classes gasoline, glue, and other neurotoxic inhalants
can produce an acute intoxication syndromeof psychoactive agents assessed in the Na-
tional Comorbidity Survey. Adjusted odds similar to that produced by alcohol, includ-
ing signs such as slurred speech, ataxia, andratios (ORs) comparing inhalant users (OR =
1.9, CI: 1.0 to 3.5), inhalant abusers (OR = stupor or coma. Onset of intoxication occurs
rapidly and may disinhibit cortical circuits10.0, CI: 2.8 to 35.0), and persons with inhal-
ant dependence (OR = 6.8, CI: 2.4 to 18.8) to regulating expression of self-injurious and
other impulsive behaviors. Thus, acute inhal-inhalant nonusers were substantial and statis-
tically significant. Inhalant dependence was ant intoxication may contribute to suicidal
ideation and/or impulsive suicide attempts onmore strongly related to suicide attempts
than any other type of substance dependence; the part of people experiencing such ideation.
Persistent neurological and cognitive impair-similar findings were noted for inhalant abuse
compared to other forms of substance abuse. ments predisposing to suicidal ideation and
suicide attempts may follow repeated use ofInhalant use without abuse or dependence
posed a significant risk for suicide attempt neurotoxins, such as that which occurs in
IUDs. It is also possible that persons experi-but at a level similar to most other forms of
substance use, suggesting that inhalant abuse encing heightened levels of depression or
anxiety may “self-medicate” with inhalants,and dependence may be relatively specific
markers of elevated risk for suicide attempts. which are reported by some users to have
anxiolytic and euphorogenic effects (Perron,In one of few longitudinal studies,
Wilcox and Anthony (2004) did not find on- Vaughn, & Howard, 2007). Finally, inhalant
use/IUDs and suicide ideation/attempts mayset of alcohol or tobacco use prior to age 16
significantly associated with later risk for sui- share common origins in factors increasing
risk for both sets of conditions, including lowcide attempt or ideation in a large sample of
youth followed from first grade into young socioeconomic status and mood, anxiety, and
personality disorders. That is, the associationadulthood. However, onset of inhalant use
prior to age 16 was a significant risk factor may be noncausal and attributable to third-
factors that increase risk for both sets of dis-for suicide attempt (Relative Risk [RR] = 2.2)
and suicide ideation (RR = 2.9) for girls, but orders.
The availability of a large, nationallynot boys. Freedenthal et al. (2007) reported
that in a sample of incarcerated adolescents, representative survey incorporating assess-
ments of inhalant use, IUDs, suicide ideationinhalant use was significantly related to sui-
cidal ideation in girls but not boys, whereas and attempts, and relevant sociodemographic
covariates enabled our team to explore rela-boys and girls with formal IUDs had signifi-
cantly elevated rates of suicide ideation and tionships among these variables in the largest
comorbidity study conducted in the Unitedattempts compared to inhalant nonusers in
their study of 723 antisocial adolescents. Sa- States. Although Borges et al.’s (2000) and
Sakai et al.’s (2004) results suggest that inhal-kai, Hall, Mikulich-Gilbertson, and Crowley
278 Suicide and Inhalants
ant users evidence greater suicidality than in- Measures
halant nonusers, it remains unclear whether
and to what extent inhalant users who prog- Face-to-face interviews were conducted
by U.S. Census Bureau workers trained byress to formal IUDs are at greater risk for
suicidal ideation and attempts than inhalant the National Institute on Alcohol Abuse and
Alcoholism. Interviewers administered theusers without IUDs. Thus, we hypothesized
that inhalant users without IUDs would evi- Alcohol Use Disorder and Associated Disa-
bilities Interview ScheduleDSM-IV versiondence significantly greater prevalence of sui-
cide ideation and attempts than inhalant (AUDADIS-IV), shown to have good reli-
ability in assessing alcohol and drug use innonusers, but significantly lower prevalence
of suicide ideation and attempts than adults the general population (Grant, Harford, Daw-
son, Chou, & Pickering, 1995).with IUDs. That is, we expected that prob-
lematic involvement with inhalants would Participants in this study included life-
time inhalant nonusers and users (i.e., re-display a graded relationship to suicide ide-
ation and attempts and that these relation- spondents who answered affirmatively to the
question, “Have you ever used inhalants orships would be stronger in women than men.
solvents, for example, amyl nitrite, nitrous
oxide, glue, toluene, or gasoline?”). An intro-
ductory statement preceding the inhalant as-
sessment item ensured that respondents were
informed that they were to report inhalants
used on their own, without a doctor’s pre-
scription, and for the purposes of getting
high, enjoying themselves, relaxing, feeling
better, feeling more alert, quieting theirStudy findings are based on data from
the 20012002 National Epidemiologic Sur- nerves, or seeing how they would work.
Within the sample of lifetime inhalant users,vey on Alcohol and Related Conditions
(NESARC). NESARC is a nationally repre- “inhalant abuse” diagnoses were assigned to
respond en ts who met DSM-IV inhalant abusesentative survey of 43,093 noninstitutional-
ized U.S. residents aged 18 years and older criteria only, and “inhalant dependence” di-
agnoses were assigned to adults who met(Grant, Dawson, Stinson, Chou, Kay, &
Pickering, 2003). The survey gathered infor- DSM-IV inhalant dependence criteria (Ameri-
can Psychiatric Association, 2000). In thismation on alcohol use and comorbid condi-
tions from individuals living in household study, respondents were classified as inhalant
nonusers if they reported no lifetime inhalantand group settings such as shelters, college
dormitories, and group homes. use, inhalant users (IUD) if they reported
lifetime inhalant use but had never metNESARC utilized a multistage cluster
sampling design, oversampling young adults, DSM-IV inhalant abuse or inhalant depen-
dence criteria, or as inhalant users with IUDsHispanics, and African Americans in the in-
terest of obtaining reliable statistical estima- (IUD+) if they reported inhalant use and met
lifetime criteria for either inhalant abuse ortion in these populations, and to ensure ap-
propriate representation of racial and ethnic dependence. Inhalant users meeting inhalant
abuse or dependence criteria were combinedsubgroups, with an overall response rate of
81%. Data were weighted at the individual into one group of respondents (IUD+)to
maximize statistical power for group con-and household level and to adjust for over-
sampling and nonresponse on demographic trasts.
NESARC participants completed sui-variables. Data were also adjusted to be rep-
resentative (viz., region, age, race, and eth- cide ideation and attempt items if they re-
sponded affirmatively to one or both of twonicity) of the U.S. adult population based on
the 2000 Census. items assessing lifetime low mood and anhe-
Howard et al. 279
donia. The questions were: “In your entire far left columns of tabled results. Adjusted
odds ratios (AORs) and 95% confidence in-life, have you ever had a time when you felt
sad, blue, or down most of the time for at tervals are presented to reflect association
strength and significance.least 2 weeks?” and “In your entire life, have
you ever had a time, lasting at least 2 weeks,
when you didn’t care about the things that
you usually cared about, or when you didn’t RESULTS
enjoy the things you usually enjoyed?” Over-
all, 13,733 (32%) respondents reported at Characteristics of Inhalant Nonusers
and IUD and IUD+ Inhalant Usersleast one lifetime episode of low mood and/
or anhedonia; approximately 66% of those
were women and the sample averaged 45.9 Comparisons of inhalant nonusers and
IUD and IUD+ respondents revealed sig-(SD = 17.2) years of age. Suicide -s pec ific ques-
tions asked whether respondents had ever at- nificant differences across measures of gen-
der, age, marital status, and lifetime DSM-IVtempted suicide, thought about committing
suicide, felt like they wanted to die, or thought diagnoses of antisocial personality disorder,
anxiety disorders, mood disorders, and alco-a lot about their own death during a period
of low mood. These questions were answered hol and non-inhalant drug use disorders (Ta-
ble 1). In general, IUD- and IUD+ userseither “yes” or “no.”
Alcohol use disorder diagnoses in- were significantly younger, more likely not to
have married, and more likely to meet anti-cluded lifetime alcohol abuse and alcohol de-
pendence and were assigned in accordance social personality, anxiety, mood, and non-
inhalant substance use disorder criteria, thanwith DSM-IV specifications. Likewise, anti-
social personality disorder, anxiety disorders inhalant nonusers.
Approximately seven-in-ten IUD re-(which included panic disorder, agoraphobia,
social and specific phobias, and generalized spondents were male, half were between ages
18 and 34, one third met antisocial personal-anxiety disorder), mood disorders (i.e., life-
time mania, hypomania, dysthymia, and ma- ity disorder criteria, half evidenced a lifetime
anxiety disorder, three quarters met lifetimejor depressive disorder), and non-inhalant
drug use disorders (i.e., lifetime abuse or de- mood disorder criteria, and more than four
in five met criteria for lifetime alcohol andpendence on heroin, hallucinogens, cocaine/
crack, marijuana, stimulants, painkillers, tran- non-inhalant drug use disorders. IUD+ re-
spondents tended to be disproportionatelyquilizers, and sedatives) were also assigned
consistent with DSM-IV guidelines. Race, male and had significantly higher rates of
lifetime antisocial personality disorder andgender, age, marital and educational status,
and annual personal income were included as non-inhalant drug use disorders and equiva-
lently high rates of alcohol use disorders andcovariates with response categories indicated
in tabled results. mood disorders compared to IUD respon-
Analytic procedures
Suicide Ideation and Attempts among
Inhalant Nonusers and IUDWeighted prevalence estimates and
95% confidence intervals were computed us- and IUD+ Respondents
ing SUDAAN Version 9.0 (Research Trian-
gle Institute, 2004). This system implements Significant group differences were
identified between inhalant-nonusing, IUD,a Taylor series linearization to adjust stan-
dard errors of estimates for complex survey and IUD+ respondents across all measures of
suicide ideation and attempt in the totalsampling design effects. Multivariate logistic
regression analyses were conducted with si- NESARC sample assessed for suicidality (Ta-
ble 2). Specifically, IUD respondents re-multaneous entry of covariates listed in the
280 Suicide and Inhalants
Characteristics of NESARC Participants Completing Suicide Items, Overall and by Inhalant
Use Subgroup
Inhalant Subgroup comparison
Overall Non-users IUD IUD+
Variable N = 13,733 N = 13,356 N = 287 N = 90 p
Male 38.09 (.52) 37.03 (.52) 69.51 (3.23) 75.82 (5.14) .0000
Female 61.91 (.52) 62.97 (.52) 30.49 (3.23) 24.18 (5.14)
Age (yr)
1834 30.36 (.55) 29.76 (.56) 51.14 (3.37) 40.64 (5.92) .0000
3554 42.87 (.52) 42.76 (.53) 43.91 (3.39) 54.69 (5.95)
55 26.77 (.52) 27.48 (.54) 4.95 (1.31) 4.67 (2.19)
White 75.66 (1.24) 75.47 (1.26) 81.88 (2.93) 80.77 (4.43) .0977
Non-White 24.34 (1.24) 24.53 (1.26) 18.12 (2.93) 19.23 (4.43)
Marital Status
Married/cohabitating 56.31 (.62) 56.77 (.64) 42.19 (3.58) 39.90 (5.85) .0000
Separated/divorced/widowed 23.21 (.43) 23.32 (.43) 18.92 (2.86) 22.29 (4.33)
Never married 20.48 (.58) 19.91 (.58) 38.17 (3.38) 37.81 (5.73)
<High school 14.69 (.46) 14.65 (.47) 14.44 (2.73) 21.01 (5.33) .7977
High school 28.03 (.62) 28.05 (.62) 27.24 (3.55) 27.92 (5.28)
>High school 57.28 (.73) 57.30 (.74) 58.32 (4.16) 51.08 (5.90)
Personal income, $/y
019,999 25.26 (.64) 25.09 (.64) 29.51 (3.27) 34.42 (6.38) .4941
20,00034,999 20.57 (.49) 20.57 (.50) 20.92 (2.66) 18.21 (4.21)
35,00069,999 31.45 (.53) 31.47 (.54) 32.14 (3.58) 26.23 (6.07)
70,000+ 22.73 (.77) 22.87 (.79) 17.43 (2.94) 21.14 (5.38)
Antisocial personality disorder
Yes 6.91 (.30) 5.87 (.27) 33.96 (3.70) 56.77 (6.24) .0000
No 93.09 (.30) 94.13 (.27) 66.04 (3.70) 43.23 (6.24)
Lifetime DSM-IV anxiety
Yes 34.80 (.69) 34.34 (.69) 50.01 (3.67) 44.63 (6.33) .0003
No 65.20 (.69) 65.66 (.69) 49.99 (3.67) 55.37 (6.33)
Lifetime DSM-IV Mood Disorder
Yes 61.4 (.66) 60.90 (.67) 74.77 (2.96) 83.83 (4.25) .0000
No 38.60 (.66) 39.10 (.67) 25.23 (2.96) 16.17 (4.25)
Lifetime DSM-IV alcohol use disorder
Yes 38.52 (.79) 36.93 (.78) 87.89 (2.63) 87.36 (4.27) .0000
No 61.48 (.79) 63.07 (.78) 12.11 (2.63) 12.64 (4.27)
Lifetime DSM-IV non-inhal-
ant drug use disorder
Yes 16.44 (.54) 14.24 (.51) 81.43 (2.83) 94.54 (2.62) .0000
No 83.56 (.54) 85.76 (.51) 18.57 (2.83) 5.46 (2.62)
Notes. *All Ns in column headings are unweighted values. All table values are weighted column
percentages (standard errors). IUD refers to inhalant users without lifetime DSM-IV inhalant abuse
or dependence. IUD+ refers to inhalant users with DSM-IV lifetime inhalant abuse or dependence.
†Tests of differences based on chi-square. In some cases, cell totals may not equal 100% due
to rounding error. Bonferroni-corrected p value = .05/10 = .005.
Pairwise post-hoc contrasts for inhalant nonusers vs. inhalant users with DSM-IV inhalant use
disorders (IUD+). All significant
contrasts were significant at p < .001.
Pairwise post-hoc contrasts for inhalant nonusers vs. inhalant users without DSM-IV inhalant
use disorders (IUD). All significant
contrasts were significant at p < .001.
Pairwise post-hoc contrasts for inhalant users without inhalant use disorders (IUD) vs. inhal-
ant users with inhalant use disorders (IUD+). The contrasts for antisocial personality disorder and
noninhalant drug use disorders were significant at p < .01.
Howard et al. 281
Proportions of NESARC Respondents with a History of Suicide Attempts and Ideation, Overall
and by Inhalant Use Subgroups
Inhalant User Subgroup Comparisons
Inhalant Non-users IUD IUD+
Total, N = 13,180 Total, N = 284 Total, N = 88
Overall Males, n = 4,379 Males, n = 185 Males, n = 63
Suicide variable N = 13,552 Females, n = 8,801 Females, n = 99 Females, n = 25 p
Suicide attempt
Total 7.48 (.28) 7.12 (.27) 18.02 (2.78) 20.15 (4.46) .0002
Males 6.49 (.46) 5.78 (.43) 18.08 (3.53) 16.73 (5.03) .0017
Females 8.08 (.33) 7.90 (.33) 17.87 (4.48) 30.85 (10.41) .0198
Thought about
committing suicide
Total 26.38 (.64) 25.68 (.62) 50.84 (3.78) 67.42 (6.17) .0000
Males 26.31 (.88) 24.81 (.87) 48.82 (4.54) 65.25 (7.19) .0000
Females 26.43 (.71) 26.19 (.70) 55.43 (5.79) 74.21 (10.02) .0001
Felt like you wanted
to die
Total 32.35 (.60) 31.69 (.59) 52.70 (3.72) 66.90 (5.45) .0000
Males 29.35 (.84) 27.87 (.83) 51.05 (4.48) 63.66 (6.91) .0000
Females 34.19 (.74) 33.93 (.73) 56.52 (5.66) 77.01 (9.21) .0003
Thought a lot about
your own death
Total 25.41 (.48) 24.92 (.47) 37.22 (3.27) 55.13 (6.49) .0000
Males 25.69 (.74) 24.90 (.75) 33.26 (3.94) 58.39 (7.53) .0009
Females 25.24 (.59) 24.93 (.58) 46.24 (5.39) 44.95 (12.07) .0011
Notes. All Ns in column headings are unweighted values. Table values are weighted column per-
centages (standard errors). IUD refers to lifetime inhalant users without lifetime DSM-IV inhalant
abuse or dependence. IUD+ refers to inhalant users with DSM-IV lifetime inhalant abuse or dependence.
†Tests of differences based on χ
. The Bonferroni-adjusted p value was .05/12 = .004.
Pairwise post-hoc contrasts for inhalant nonusers vs. inhalant users with DSM-IV inhalant use
disorders (IUD+). For total sample, all
contrasts were significant at p < .001, except suicide attempt (p
< .01). For males, thought about committing suicide and felt like you wanted to die were significant at
p < .001, thought a lot about your own death at p < .01, and suicide attempt at p < .05. For females,
thought about committing suicide and felt like you wanted to die were significant at p < .01, and suicide
attempt was significant at p < .05.
Pairwise post-hoc contrasts for inhalant nonusers vs. inhalant users without DSM-IV inhalant
use disorders (IUD). All such contrasts for the total sample were significant at p < .001. For men,
thought about committing suicide and felt like you wanted to die were significant at p < .001, suicide
attempt at p < .01, and thought a lot about your own death at p < .05. For females, thought about com-
mitting suicide and thought a lot about your own death were significant at p < .001, felt like you wanted
to die at p < .01, and suicide attempt at p < .05.
Pairwise post-hoc contrasts for inhalant users without inhalant use disorders (IUD) vs. inhalant
users with inhalant use disorders (IUD+). All significant
contrasts for the total sample were significant
at p < .05. For males, thought a lot about your own death was significant at p = .01. For females, no post
hoc contrasts were statistically significant; p values ranged from .10 to .22 for the variables in the first
three rows of Table 2, suggesting that low statistical power may have been a factor in the failure to
detect significant IUD vs. IUD+ in women across measure of suicide despite apparently moderate-to-
large effects.
282 Suicide and Inhalants
ported significantly higher lifetime preva- Other factors associated with suicide
ideation and attempts included being female;lence rates of thinking about suicide, feeling
like they wanted to die, and thinking a lot younger; separated; divorced, or widowed;
currently married or cohabitating (inverselyabout their own death than inhalant nonus-
ers, but significantly lower rates of suicide associated); having lower annual incomes and
less education, and having met criteria forideation compared to IUD+ respondents.
The prevalence of suicide ideation was very antisocial personality, anxiety, mood, alcohol
use, and non-inhalant drug use disorders. Ithigh (55%-67%) for IUD+ respondents
across all three measures. IUD+ women re- is notable that even with the effects of a sig-
nificant number of psychiatric, substance-ported higher rates of thinking about com-
mitting suicide and feeling like they wanted related, and sociodemographic variables con-
trolled for, IUD+ adults were at significantlyto die than IUD+ men, whereas IUD+ men
were more likely to report thinking a lot elevated risk for suicidal ideation.
Odds ratios reflecting comparisons ofabout their own death than IUD+ women.
IUD and IUD+ respondents exhib- IUD to IUD+ inhalant users across mea-
sures of lifetime suicide attempt, thinkingited significantly elevated rates of one or
more lifetime suicide attempts compared to about committing suicide, feeling like you
wanted to die, and thinking about your owninhalant nonusers, although differences be-
tween IUD and IUD+ respondents were death were not significant (Table 3; analyses
available upon request).not substantial overall or for men. Nearly
one-in-three IUD+ women reported a life-
time suicide attempt compared to fewer than
one-in-five IUD women and one-in-thir- DISCUSSION
teen women who were inhalant nonusers, al-
though these differences did not attain statis-
Study findings provide partial support
for a positive association between severity of
tical significance due to low power. Thus, for
problematic inhalant involvement and suicid-
all three measures of suicide ideation among
ality. In bivariate analyses, prevalence of sui-
women, there was bivariate evidence of an
cide ideation in inhalant users without IUDs
monotonic association between extent of
was much higher than that reported by inhal-
problematic involvement with inhalants and
ant nonusers, but lower than the prevalence
suicidal thoughts. Inhalant users evidenced
of suicide ideation reported by adults with
significantly higher rates of suicide attempts
IUDs. Multivariate logistic regression analyses,
than did nonusers.
controlling for psychiatric, substance-related,
and sociodemographic factors, indicated thatMultivariate Logistic Regression
Prediction of Suicide Ideation IUD+ respondents had substantially greater
odds of reporting suicidal ideation than in-and Attempts
halant nonusers.
Bivariate analyses revealed a signifi-Mu lt iva ri ate logisti c regress io n was used
to examine associations between inhalant use cantly greater prevalence of lifetime suicide
attempt s among IUD respond ent s comparedand each suicide ideation and attempt mea-
sure, while adjusting for other sociodemo- to inhalant nonusers and indicated that that
IUD men and women reported similar pro-graphic and clinical variables. The results in-
dicated that IUD+ but not IUD respondents portions with lifetime suicide attempts, but
IUD+ women reported nearly twice thewere significantly more likely to report sui-
cidal ideation compared to inhalant nonusers. prevalence of suicide attempts as IUD+ men
(and IUD men and women). In multivariateNeither IUD or IUD+ respondents exhib-
ited a significantly higher likelihood of re- logistic regression analyses, women were sig-
nificantly more likely then men to attemptporting actual suicide attempts compared to
inhalant nonusers. suicide but no significant main effects on sui-
Howard et al. 283
Adjusted Odds Ratios Reflecting Associations Between Inhalant Use and Suicide Attempts and Ideation
Among NESARC Respondents Completing Suicide Attempt and Ideation Items, N = 13,552
Suicide Think about Feel like wanted Think about
Independent attempt committing suicide to die own death
Variables AOR (95% CI) AOR (95% CI) AOR (95% CI) AOR (95% CI)
Inhalant group
IUD 1.23 (.801.90) 1.40 (.962.05) 1.38 (.961.99) .93 (.671.30)
IUD+ 1.09 (.582.06) 2.19 (1.184.05) 1.99 (1.183.38) 1.62 (.912.86)
Male .67 (.55.81) .90 (.811.01) .76 (.69.85) 1.01 (.911.14)
Age (yr)
35–54 .77 (.62.95) .89 (.771.02) .95 (.841.08) .87 (.77.99)
55 .47 (.36.62) .73 (.62 .87) .85 (.721.0) .80 (.69.92)
Non-White 1.18 (.991.41) .89 (.781.01) 1.05 (.931.17) 1.22 (1.091.36)
Marital status
Never married†
Married/cohabitating .91 (.721.15) .81 (.71 .93) .86 (.75 .98) .79 (.70 .90)
widowed 1.43 (1.131.80) 1.02 (.861.22) 1.07 (.911.26) .89 (.771.02)
>High school†
High school 1.09 (.901.32) .91 (.811.03) 1.13 (1.011.26) 1.00 (.891.13)
<High school 1.38 (1.081.77) .94 (.811.09) 1.20 (1.031.39) 1.11 (.951.31)
Personal income, $/y
20,00034,999 .75 (.60.92) .88 (.761.02) .88 (.761.02) .94 (.821.08)
35,00069,999 .63 (.51.79) .83 (.72 .97) .79 (.68 .91) .86 (.75 .99)
70,000+ .51 (.37.70) .72 (.61 .86) .71 (.60 .85) .87 (.721.05)
Antisocial personality
Yes 2.32 (1.832.96) 2.02 (1.622.52) 1.58 (1.311.91) 1.72 (1.392.13)
Lifetime anxiety disorder
Yes 1.41 (1.181.69) 1.39 (1.251.54) 1.45 (1.321.59) 1.49 (1.351.65)
Lifetime alcohol use
Yes 1.30 (1.071.58) 1.31 (1.161.48) 1.23 (1.101.37) 1.18 (1.051.33)
Lifetime mood disorder
Yes 5.43 (4.117.19) 6.15 (5.337.09) 7.22 (6.328.25) 4.15 (3.634.74)
Lifetime non-inhalant drug
use disorder
Yes 1.56 (1.261.93) 1.51 (1.321.72) 1.41 (1.231.62) 1.26 (1.081.47)
Notes. Values in bold are significant based on a 95% confidence interval that does not bound 1.0.
All reported odds ratios were adjusted for the remaining variables listed in the far left column of the
AOR = Adjusted odds ratio.
IUD refers to lifetime inhalant users without lifetime DSM-IV inhalant abuse or dependence.
IUD+ refers to lifetime inhalant users with DSM-IV lifetime inhalant abuse or dependence.
= reference group.
284 Suicide and Inhalants
cide attempts were observed for status as a of suicidal behavior, particularly suicide at-
tempts.IUD or IUD+ inhalant user relative to in-
halant nonuser. As noted above, post hoc mul- Particularly striking in the current
study were the findings that nearly one-in-tiple logistic regression follow-up analyses
testing gender times inhalant-user-category three IUD+ women had attempted suicide
and that more than half of IUD and threeinteraction terms across lifetime suicide at-
tempts were nonsignificant. quarters of IUD+ women had thought about
committing suicide and felt like they wantedThus, our findings support a significant
relationship between inhalant use, IUDs, and to die, respectively. These results are com-
patible with contemporary reports indicatingsuicide ideation, but provide less clear find-
ings with respect to suicide attempts. IUD that girls and women who use inhalants and/
or develop IUDs are at elevated risk for sui-and IUD+ adults had lifetime suicide attempt
prevalence rates nearly three times greater cidal thoughts and attempts. For example,
Luncheon et al. (2008) found high rates ofthan inhalant nonusers in bivariate analyes,
but inhalant use and IUDs were not related inhalant use in adolescent females contem-
plating suicide and an odds ratio of 2.6 forto suicide attempt in multivariate analyses. It
is possible that the null findings for inhalant suicidal thoughts among adolescent females
who had versus had not used inhalants. Theuse and suicide attempts in multivariate anal-
yses were due to the statistical procedures investigations of Freedenthal, Vaughn, Jen-
son, and Howard (2007) and Wilcox and An-employed, which controlled comprehensively
for the effects of psychiatric comorbidities thony (2004) provide further indication that
inhalant use in females may be more stronglycommonly associated with suicide and occur-
ring disproportionately in IUD and IUD+ associated with suicidality than in males.
Possible reasons for a significant asso-adults compared to inhalant nonusers ADDIN{
ADDIN EN.CITE.DATA (Wu & Howard, ciation of inhalant use to suicidality among
girls and women are that inhalant use in fe-2007; Wu, Howard, & Pilowsky, 2008). An-
other possibility is that IUD and especially males may reflect a comparatively more psy-
chiatrically troubled or antisocial profile rela-IUD+ individuals tend to use more lethal
means for attempting suicide than non-inhal- tive to such use in male inhalant users; may
be attributable to self-medication with inhal-ant users and were therefore not included in
the NESARC survey if they completed sui- ants of anxious and depressive mood states
that are observed more frequently in femalescide. Testing this hypothesis would require
data on completed suicides among nonusers than males (Perron et al., 2007); or that in-
halant use may disinhibit suicide ideation andand IUD and IUD+ individuals, which are
currently unavailable. attempts in girls and women with low mood,
dysthymia, and major depressive disorder,The causal role of inhalant use and
IUDs in suicide attempts remains unclear which they experience at disproportionately
high levels (Howard, Balster, Cottler, Wu, &and continued study of these associations is
clearly needed. Further, Borges et al. (2000) Vaughn, 2008; Wu et al., 2008). At a mini-
mum, the findings suggest that treatment andreported disaggregated findings which sug-
gested that the effects of substance use may prevention resources should be targeted to
girls and women using inhalants.primarily be on suicide ideation and un-
planned suicide attempts, rather than on With regard to general profiles, it is
notable that among persons with a history ofplanned attempts. Mann (2003) has noted the
role that substance-related neurotoxicity, low mood, IUD+ adults tend to be compara-
tively young, unmarried, antisocial, and af-particularly in the ventral prefrontal cortex,
may play in disinhibiting aggressive behavior flicted with high rates of anxiety, mood, alco-
hol, and non-inhalant drug use disordersand promoting impulsive acts such as suicide
attempts. Thus, greater specificity would be compared to inhalant nonusers. While IUD+
women evidenced high levels of suicide at-desirable in the assessment of specific forms
Howard et al. 285
tempt and ideation, it is also true that nearly Future investigations should include
more extensive assessments of use and abuse/two thirds of IUD+ men report having
thought about committing suicide and feel- dependence on specific inhalant products and
should focus on pathogenetic processes thating like they wanted to die. The high risk
profile that characterizes IUD+ adults (and contribute to findings of significant associa-
tions between suicidality and inhalant use,IUD adults to a lesser though still signifi-
cant degree) has been reported in other in- particularly among girls and women. Find-
ings of this report are consistent with recentvestigations of youth and adult inhalant users
(Howard et al., 2008; Wu & Howard, 2007; studies suggesting that inhalant use is associ-
ated with completed suicide (Wick, Gilbert,Wu et al., 2008).
This research is characterized by sev- Felgate, & Byard, 2007) and suicide attempts
in populations ranging from schizophrenicseral limitations. Although data were collected
via a structured psychiatric interview using a (Shoval et al., 2006) to homeless and runaway
youth (Greene & Ringwalt, 1996) to menpsychometrically sound instrument, they are
based on self-report and subject to typical who have sex with men (Botnick, Heath,
Cornelisse, Strathdee, Martindale, & Hogg,limitations of self-report data including un-
reliable retrospective recall. Further, cross- 2002). Far more research is needed to ad-
dress the potential role that inhalant use andsectional studies cannot establish causality.
Thus, the etiological nature of the relation- IUDs, which are prevalent in many nations,
may play in suicidal ideation, suicide at-ship among inhalant use, IUDs, and suicidal-
ity cannot be determined in the present study. tempts, and completed suicides.
American Psychiatric Association. (2000). D. A., Chou, P. S., & Pickering, R. P. (1995).
The Alcohol Use Disorder and Associated Disa-Diagnostic and statistical manual of mental disorders
(4th ed.). Washington, DC: Author. bilities Interview Schedule (AUDADIS): Reliabil-
ity of alcohol and drug modules in a general popu-Bertolote, J. M., Fleischmann, A., De
Leo, D., & Wasserman, D. (2004). Psychiatric di- lation sample. Drug and Alcohol Dependence, 39,
3744.agnoses and suicide: Revisiting the evidence. Cri-
sis, 25, 147155. Greene, J. M., & Ringwalt, C. L. (1996).
Youth and familial substance use’s association withBorges, G., Walters, E. E., & Kessler,
R. C. (2000). Associations of substance use, abuse, suicide attempts among runaway and homeless
youth. Substance Use & Misuse, 31, 10411058.and dependen ce with subseque nt suicidal behavio r.
American Journal of Epidemiolog y , 151, 781–789. Howard, M. O., Balster, R. L., Cottler,
L. B., Wu, L.-T., & Vaughn, M. G. (2008). In-Botnick, M. R., Heath, K. V., Cornel-
isse, P. G., Strathdee, S. A., Martindale, S. L., halant use among incarcerated adolescents in the
United States: Prevalence, characteristics, and& Hogg, R. S. (2002). Correlates of suicide at-
tempts in an open cohort of young men who have correlates of use. Drug and Alcohol Dependence, 93, with men. Canadian Journal of Public Health,
93, 5962. Luncheon, C., Bae, S., Gonzalez, A.,
Lurie, S., & Singh, K. P. (2008). Hispanic femaleFreedenthal, S., Vaughn, M. G., Jenson,
J. M., & Howard, M. O. (2007). Inhalant use and adolescents’ use of illicit drugs and the risk of sui-
cidal thoughts. American Journal of Health Behav-suicidality among incarcerated youth. Drug and
Alcohol Dependence, 90, 8188. ior, 32, 5259.
Mann, J. J. (2003). Neurobiology of sui-Grant, B. F., Dawson, D. A., Stinson,
F. S., Chou, P. S., Kay, W., & Pickering, R. cidal behavior. Nature Reviews, 4, 819828.
Perron, B. E., Vaughn, M. G., & How-(2003). The Alcohol Use Disorder and Associated
Disabilities Interview Schedule-IV (AUDADIS- ard, M. O. (2008). Reasons for using inhalants:
Evidence for discrete classes in a sample of incar-IV): Reliability of alcohol consumption, tobacco
use, family history of depression and psychiatric cerated adolescents. Journal of Substance Abuse
Treatment, 34(4), 450455.diagnostic modules in a general population sam-
ple. Drug and Alcohol Dependence, 71, 716. Sakai, J. T., Hall, S. K., Mikulich-
Gilbertson, S. K., & Crowley, T. J. (2004). In-Grant, B. F., Harford, T. C., Dawson,
286 Suicide and Inhalants
halant use, abuse, and dependence among adoles- Wilcox, H. C., Conner, K. R., & Caine,
E. D. (2004). Association of alcohol and drug usecent patients: Commonly comorbid problems.
Journal of the American Academy of Child and Ado- disorders and completed suicide: An empirical re-
view of cohort studies. Drug and Alcohol Depen-lescent Psychiatry, 43, 10801088.
Shoval, G., Sever, J., Sher, L., Diller, R., dence, 76 (suppl.) S11S19.
Wu, L.-T., & Howard, M. O. (2007).Apter, A., Weizman, A., et al. (2006). Substance
use, suicidality, and adolescent-onset schizophre- Psychiatric disorders in inhalant users: Results
from the National Epidemiologic Survey on Alco-nia: An Israeli 10-year retrospective study. Journal
of Child and Adolescent Psychopharmacology, 16, 767 hol and Related Conditions. Drug and Alcohol De-
pendence, 88, 146155.775.
Wick, R., Gilbert, J. D., Felgate, P., & Wu, L. T., Howard, M. O., & Pilowsky,
D. J. (2008). Substance use disorders among inhal-Byard, R. W. (2007). Inhalant deaths in South
Australia: A 20-year retrospective autopsy study. ant users: Results from the National Epidemio-
logic Survey on Alcohol and Related Conditions.American Journal of Forensic Medicine and Pathology,
28, 319322. Addictive Behaviors, 33, 968973.
Wilcox, H. C., & Anthony, J. C. (2004).
The development of suicide ideation and at- Manuscript Received: June 7, 2009
Revision Accepted: June 7, 2009tempts: An epidemiologic study of first graders
followed into young adulthood. Drug and Alcohol
Dependence, 76 (suppl.) S53S67.
  • [Show abstract] [Hide abstract] ABSTRACT: Anecdotal reports and recent medical examiner and poison control center studies suggest that computer duster spray (CDS) inhalation is an emerging public health threat. However, there is a current dearth of empirical data on CDS use. Study aims were to examine the prevalence, frequency, correlates, and modalities of CDS use among a treatment sample of antisocial youth. Methods: A battery of standardized psychosocial instruments was administered via interview of 723 Missouri adolescents in residential care for antisocial behavior. Lifetime CDS use was prevalent (14.7%) in this young service population (97.7% of whom participated). CDS users were significantly more likely to report histories of perinatal injuries or illness, traumatic experiences, suicidality and physician-diagnosed mental illness, and evidenced higher levels of psychiatric symptoms, antisocial attitudes and behaviors, and polydrug use than CDS nonusers. CDS use was endemic in this treatment sample of adolescents and associated with a range of clinically significant comorbidities. Current findings describe an underrecognized and potentially dangerous form of substance misuse that has rarely been studied but that may be of growing importance.
    Article · Nov 2010
  • [Show abstract] [Hide abstract] ABSTRACT: To address gaps identified in earlier reviews, namely, the comparative influence of substance use on attempted suicide and completed suicide; the proximal role of substance use as a trigger of suicidal behaviour; the association between substances and suicidal behaviour; suicide and substance use disorders in youth; and the influence of combinations of risk factors, including psychiatric morbidity, on suicide. The presence of an alcohol use disorder is confirmed as a distal risk factor for completed suicide, as well as attempted suicide. Alcohol use at the time of the suicide attempt is associated with low-risk methods. The use of other substances as a trigger of suicidal behaviour is highlighted in recent studies, but the circumstances leading to the suicidal act and the direct influence of substances in suicidal behaviour need to be explored further. Inhalant use and cocaine use are particularly associated with suicidal behaviour. Young people with multiple risk behaviours, such as substance use and risky sexual behaviours, are at high risk for suicidal behaviour. Psychiatric comorbidity with substance use escalates the risk for suicidal behaviour. Environmental interventions, such as reduction in the number of bars, may be helpful in reducing alcohol-related morbidity, including suicides. Prevention strategies for reducing suicidal behaviour among substance users need to be prioritized at the individual and the national level through effective design, policy and implementation.
    Full-text · Article · Mar 2011
  • [Show abstract] [Hide abstract] ABSTRACT: Inhalants are widely misused by adolescents and are among the most toxic of psychoactive substances. This investigation examined the prevalence and correlates of adverse consequences of acute inhalant intoxication. Adolescent inhalant users (n = 279) in residential care completed structured interviews including assessments of the characteristics of their inhalant use. Multivariate logistic and linear regression and path analyses identified correlates of adverse inhalant intoxication-related experiences. Results of this study indicated that high-risk behaviors and adverse outcomes experienced during episodes of inhalant intoxication were common in this sample. High-frequency inhalant users were significantly more likely than moderate- and low-frequency users to experience adverse consequences of inhalant intoxication. Certain risky behaviors and consequences, such as engaging in unprotected sex or acts of physical violence while high on inhalants, were dramatically more common among high-frequency users than low-frequency users. Prior traumatic experiences, trait impulsivity, self-medication use of inhalants, and polydrug use were significant correlates of adverse inhalant-intoxication-related consequences. Adverse events and high-risk behaviors commonly occurred during episodes of inhalant intoxication in this sample of adolescents. High-frequency inhalant users and youth who used inhalants to medicate negative affective states were at elevated risk for such events.
    Article · Apr 2011
Show more