Association Between Exposure to Suicidal Behaviors and Suicide Attempts Among Adolescents: The Moderating Role of Prior Psychiatric Disorders

Abstract

Theories suggest that adolescents exposed to suicide attempts and/or deaths are at higher risk of attempting suicide, themselves. However, research findings have been mixed, with most studies not accounting for psychiatric vulnerability. We examined past psychiatric diagnosis as a moderator of the association between lifetime exposure to suicide attempts and/or deaths and adolescents’ suicide attempts. Adolescents (N = 518; 60% female; 45% White), ages 12–21, reported on prior suicide ideation and attempts, and mood, anxiety, and substance use disorders at baseline. Suicide attempts since baseline and exposure to suicidal behaviors were assessed 4–6 years later. Lifetime exposure to family suicide attempts and/or suicide deaths, but not to suicidal behaviors of peers/friends or others, was associated with a suicide attempt at follow-up among those with prior psychiatric disorders. Psychologically vulnerable adolescents may require additional support after exposure to suicidal behaviors of a family member to reduce their risk of attempting suicide.

Full text

Vol.:(0123456789)
1 3
Child Psychiatry & Human Development (2022) 53:365–374
https://doi.org/10.1007/s10578-021-01129-2
ORIGINAL ARTICLE
Association Between Exposure toSuicidal Behaviors andSuicide
Attempts Among Adolescents: The Moderating Role ofPrior
Psychiatric Disorders
EmilyA.Kline1,2· AnaOrtin‑Peralta1,3· LillianPolanco‑Roman4,5· ReginaMiranda1,6,7
Accepted: 17 January 2021 / Published online: 10 February 2021
© The Author(s), under exclusive licence to Springer Science+Business Media, LLC part of Springer Nature 2021
Abstract
Theories suggest that adolescents exposed to suicide attempts and/or deaths are at higher risk of attempting suicide, them-
selves. However, research findings have been mixed, with most studies not accounting for psychiatric vulnerability. We
examined past psychiatric diagnosis as a moderator of the association between lifetime exposure to suicide attempts and/
or deaths and adolescents’ suicide attempts. Adolescents (N = 518; 60% female; 45% White), ages 12–21, reported on prior
suicide ideation and attempts, and mood, anxiety, and substance use disorders at baseline. Suicide attempts since baseline
and exposure to suicidal behaviors were assessed 4–6years later. Lifetime exposure to family suicide attempts and/or suicide
deaths, but not to suicidal behaviors of peers/friends or others, was associated with a suicide attempt at follow-up among
those with prior psychiatric disorders. Psychologically vulnerable adolescents may require additional support after exposure
to suicidal behaviors of a family member to reduce their risk of attempting suicide.
Keywords Suicide· Suicide attempt· Exposure· Adolescents· Psychiatric disorders
Adolescent suicide is a major public health concern in the
U.S. and worldwide [1]. In the U.S., adolescent suicide
rates have nearly tripled over the last decade [2]. Increases
in adolescent suicide deaths (SDs) further suggest greater
exposure to suicidal behaviors in young populations, and
knowing someone who has attempted or died by suicide
may increase risk for suicidal behaviors among adolescents
[3, 4]. Moreover, the incidence of suicide attempts (SAs)
surges during adolescence [5] and increases in lethality
through young adulthood [6]. Identifying early indicators
of suicide-related risk in adolescence is critical, and a better
understanding of the harmful effects of exposure to suicidal
behaviors may help address this need.
Exposure to fatal or non-fatal suicidal behaviors (i.e.,
deaths, attempts) is associated with elevated risk of adoles-
cent suicidal behavior [3, 4]. This elevated risk is evident
across various avenues, including more direct exposure
through family [7–12], friends [13, 14], and high school
peers [15], or indirect exposures through the media [16, 17].
Imitation or modeling behavior drawn from a social learning
framework is often cited as a potential explanation for this
elevated risk [3, 4]. This explanation is also consistent with
the Interpersonal Theory of Suicide [18], which highlights
the role of social connections (i.e., thwarted belongingness,
perceived burdensomeness) in suicide risk. This model
emphasizes that an individual can acquire the capability for
suicide via a reduced fear threshold that may result from
exposure to suicidal behaviors. Being exposed to suicidal
behaviors, thus, may impact risk for SAs among adolescents
through intersecting intrapersonal, interpersonal, and envi-
ronmental influences.
* Regina Miranda
regina.miranda@hunter.cuny.edu
1 Hunter College, City University ofNew York, NewYork,
USA
2 Montclair State University, Montclair, USA
3 Ferkauf Graduate School ofPsychology, Yeshiva University,
NewYork, USA
4 Columbia University Medical Center, New York State
Psychiatric Institute, NewYork, USA
5 The New School forSocial Research, NewYork, USA
6 The Graduate Center, City University ofNew York,
NewYork, USA
7 Department ofPsychology, Hunter College, City University
ofNew York, 695 Park Ave., Room 611HN, NewYork,
NY10065, USA

366 Child Psychiatry & Human Development (2022) 53:365–374
1 3
One of these intrapersonal factors is the presence of pre-
existing psychiatric disorders. Mood, anxiety, and substance
use disorders have been identified as the main underlying
conditions among adolescents who died by suicide [19, 20]
and are associated with suicidal behaviors among adoles-
cents [5, 21]. Few studies have addressed whether knowing
someone who has died by suicide most impacts adolescents
with pre-existing psychiatric symptoms, with mixed findings
[22–24]. One cross-sectional study of 12–17-year-olds in
Hong Kong found that the association between exposure to
SAs, but not SDs, by someone they knew and adolescents’
risk for attempting suicide was stronger among those with
elevated depressive symptoms [24]. A study of 7th–12th
-graders in the U.S. indicated that severity of depressive
symptoms did not impact the relation between past-year
exposure to SAs of family or peers and a past-year SA
among adolescents [22]. Similarly, a longitudinal study of
Canadian adolescents found that exposure to a schoolmate’s
SD, assessed at ages 14–15, predicted an adolescent’s SA at
ages 16–17, and that past-week psychiatric symptoms (i.e.,
depression, anxiety) did not influence this relation [23].
These studies used self-report measures or a checklist that
assessed symptoms over a prior 1-week period. As such,
it is unclear whether these measures were able to capture
longer-lasting psychiatric conditions. Additional research
is needed to understand the effect of exposure to suicidal
behavior among vulnerable adolescents, in particular those
with psychiatric disorders linked to heightened suicide risk.
Studies have examined the impact of exposure based on
the relationship to the adolescent (e.g., family, peers), and
also the impact of cumulative exposure to suicidal behav-
iors—i.e., of exposure to more than one SA or SD—with
mixed results. Cross-sectional research found that exposure
to the SA of a family member, but not of a peer, was asso-
ciated with an adolescent’s risk of endorsing a SA in the
previous 12months [25], whereas others found that expo-
sure to family and friends’ SAs and SDs increased the risk
for attempting suicide when the behaviors were examined
together [22] or separately [26]. Finally, one longitudinal
study of adolescent girls in the U.S. found that exposure to a
family member’s SA was not associated with an adolescent’s
SA 2 or 6 years later. However, exposure to a peer’s SA was
associated with an adolescent’s SA 2, but not 6, years later
[13]. Only two studies of which we are aware have examined
the effect of cumulative exposure. A study of offspring of
depressed parents found that neither the number of expo-
sures to SDs or SAs (i.e., cumulative exposure was meas-
ured continuously) nor the relation to decedent or attempters
were associated with SAs among youth over age 10 [7]. In
contrast, a later study of a population-based birth cohort
found that adolescents with exposure to self-injury (with
suicidal or non-suicidal intent) by both family and friends
had a higher risk of attempting suicide at age 16 than those
exposed only to self-injury by either family or friends [21].
Taken together, this evidence from cross-sectional and lon-
gitudinal research is equivocal about whether exposure to
the SA or SD of a family member, peer, or both, differen-
tially impact risk of an adolescent’s own SA, and also about
whether cumulative exposure increases risk.
Given previous mixed findings, the present study sought
to examine the association between exposure to suicidal
behaviors and risk for attempting suicide among a racially/
ethnically diverse sample of adolescents followed up into
emerging adulthood. First, we examined how the associa-
tion between exposure to suicidal behaviors and adolescents’
SAs varied by the presence of prior psychiatric disorders. We
hypothesized that exposure to a SA and/or a SD would be
significantly related to adolescents’ SA, and that this relation
would be stronger among adolescents with (versus without)
a prior mood, anxiety, or substance use disorder. Second,
we examined exposure based on the relationship to the ado-
lescent (i.e., family, peers/friends, and others—a category
that has often been excluded from prior studies) on risk for
attempting suicide among adolescents with and without psy-
chiatric disorders, separately. Finally, we conducted explora-
tory analyses to examine the relationship between cumulative
exposure to suicidal behaviors and risk of attempting suicide.
Methods
Sample
Participants were 518 adolescents (60% female), ages 12–21
(M = 15.6, SD = 1.4), recruited from seven high schools
(public, parochial, and vocational) in the New York City
metropolitan area from 1991 to 1994, as part of a two-stage
screening, and who took part in a follow-up assessment
4–6years later (see Table1 for sample characteristics).
At baseline, 1729 adolescents completed the Columbia
Suicide Screen (CSS) [27], which assessed for prior suicide
ideation and lifetime SA. Of these adolescents, 641 ado-
lescents also completed the mood, anxiety, and substance
use modules of the Computerized Diagnostic Interview
Schedule for Children (C-DISC version 2.3), given the rel-
evance of these disorders to youth suicide [20]. Adolescents
were selected to complete the C-DISC based on whether
they screened positive for suicide risk on the CSS, such that
two-thirds of the sample had endorsed either suicide idea-
tion, a SA, or emotion-related symptoms, while one-third of
the sample (matched by grade, sex, and race/ethnicity) did
not [27]. Adolescents (N = 552; 86%) were re-contacted by
telephone 4–6years later to answer questions about expo-
sure to a SA or a SD and whether they had made a SA since
baseline. Thirty-four adolescents were excluded from this
study due to lack of information about their exposure to SAs

367Child Psychiatry & Human Development (2022) 53:365–374
1 3
or SDs. There were no significant differences between the
baseline and final sample in age or sex. However, there was
a difference by race/ethnicity,
𝜒2
= 9.99, p < .05, with fewer
Hispanic adolescents in the follow-up sample (17%) than in
the baseline sample (28 %), Z = 2.5, p < 0.05.
Table 1 Association between the demographic and mental health variables and having attempted suicide at follow-up
Bolded values are statistically significant at the p<.05 level
* Suicide ideation was excluded from the calculation of total mood symptoms and overall symptoms, in order to reduce overlap with suicide idea-
tion, as measured by the Columbia Suicide Screen
Total sample
N = 518
No SA at follow-up
N = 478
SA at follow-up
N = 40
SA vs. No SA
at follow-up
p
N (%) N (%) N (%)
𝜒2
Race/ethnicity 4.31 0.37
White 231 (44.6) 219 (45.8) 12 (30)
Black 130 (25.1) 118 (24.7) 12 (30)
Hispanic 89 (17.2) 79 (16.5) 10 (25)
Asian 39 (7.5) 36 (7.5) 3 (7.5)
Other 29 (5.6) 26 (5.4) 3 (7.5)
Sex
Female 312 (60.2) 283 (59.2) 29 (72.5) 2.72 0.10
Male 206 (39.8) 195 (40.8) 11 (27.5)
Prior suicide ideation 172 (33.2) 146 (30.5) 26 (65.0) 19.76 < 0.01
Lifetime SA 78 (15.1) 58 (12.1) 20 (50.0) 41.38 < 0.01
Prior psychiatric disorders
Mood 66 (12.7) 54 (11.3) 12 (30.0) 11.61 < 0.01
Anxiety 95 (18.3) 80 (16.7) 15 (37.5) 10.63 < 0.01
Substance use 26 (5.0) 21 (4.4) 5 (12.5) 5.09 0.02
All 130 (25.1) 109 (22.8) 21 (52.5) 17.32 < 0.01
Type of lifetime exposure
SA 223 (43.1) 202 (42.3) 21 (52.5) 1.58 0.21
SD 156 (30.1) 142 (29.7) 14 (35.0) 0.49 0.48
SA/SD 300 (57.9) 272 (56.9) 28 (70.0) 2.60 0.11
Lifetime exposure based on the
relationship to the adolescent
SA
Family 34 (6.6) 28 (5.9) 6 (15) 5.03 0.03
Peers/friends 190 (36.7) 174 (36.4) 16 (40) 0.21 0.65
Others 21 (4.1) 19 (4.0) 2 (5) 0.10 0.75
SD
Family 25 (4.8) 21 (4.4) 4 (10) 2.53 0.11
Peers/friends 119 (23.0) 109 (22.8) 10 (25) 0.10 0.75
Others 47 (9.1) 45 (9.4) 2 (5) 0.87 0.35
SA/SD
Family 56 (10.8) 46 (9.6) 10 (25) 9.05 < 0.01
Peers/friends 254 (49.0) 231 (48.3) 23 (57.5) 1.24 0.27
Others 64 (12.4) 61 (12.8) 3 (7.5) 0.94 0.33
M (SD) t p
Total symptoms
Mood*1.49 (3.53) 1.31 (3.33) 3.92 (4.87) 3.16 < 0.01
Anxiety 5.32 (7.12) 4.99 (6.77) 9.25 (9.72) 2.72 0.01
Substance use 0.62 (2.07) 0.61 (2.09) 0.70 (1.81) 0.24 0.81
All*7.37 (10.26) 6.86 (9.67) 13.89 (14.79) 2.80
Cumulative exposure to SA/SD 1.34 (2.05) 1.29 (1.90) 1.90 (3.29) 1.81 0.07

368 Child Psychiatry & Human Development (2022) 53:365–374
1 3
Procedures
Information about the study was sent home to parents, who
were given the option to opt-out of their child’s participa-
tion. Adolescents whose parents did not opt them out of
the study provided written informed assent in class [27].
The Institutional Review Boards (IRBs) of the New York
State Psychiatric Institute, the New York State Board of
Education, and the Archdiocese of New York approved the
study procedures, and the current analyses were approved
by the IRBs of the City University of New York and
Montclair State University.
Measures
Lifetime Suicide Attempts andPrior Suicide Ideation
At baseline, adolescents completed the Columbia Suicide
Screen (CSS) [27], which inquired about lifetime SAs
(“Have you ever tried to kill yourself?”) and prior suicide
ideation (“During the past 3 months, have you thought
about killing yourself?”), within a larger survey assess-
ing emotion-related symptoms (e.g., irritability, anxiety,
substance use) and health problems (e.g., headaches, dizzi-
ness). The test–retest reliability of the CSS has been found
to be 0.58 (κ) for lifetime SAs and 0.48 (κ) for suicide
ideation in the previous 3 months [27]. To assess SAs at
follow-up, adolescents completed a modified version of
the Adolescent Suicide Interview [28], a semi-structured
interview obtaining information about an adolescent’s SA
history. The follow-up interview inquired about whether
adolescents attempted suicide since the baseline assess-
ment and the characteristics of each SA made in their
lifetime.
Lifetime Exposure toSuicide Attempts and/orSuicide
Deaths
At follow-up, adolescents reported on how many people they
knew that attempted and/or died by suicide, the person’s sex,
and their relationship to the person. Lifetime exposure was
coded categorically as knowing someone who had either
attempted and/or died by suicide (“SA/SD”). This variable
was broken down by relationship to the adolescent: family
(i.e., parents, siblings, aunts, uncles, grandparents, cousins,
and any other extended family members), peers/friends (i.e.,
friends, classmates, romantic partners, acquaintances around
the same age), and others (i.e., any adult who did not fit
into the other categories). The total number of people that
the adolescent knew who attempted and/or died by suicide
reflected cumulative exposure to a SA/SD.
Prior Psychiatric Disorders
Mood, anxiety, and substance use disorders in the previous
6 months were assessed at baseline with the Computerized
Diagnostic Interview Schedule for Children, version 2.3
(C-DISC-2.3) [29], a structured interview that assessed psy-
chiatric disorders according to DSM-III-R criteria and was
administered by lay interviewers. Specifically, the psychiatric
disorders examined included: mood (i.e., major depressive
disorder, dysthymic disorder); anxiety (i.e., panic disorder,
agoraphobia, social phobia, generalized anxiety disorder,
overanxious disorder); and substance use (i.e., alcohol abuse/
dependence, marijuana abuse/dependence, other substance
use/dependence). While we were primarily interested in exam-
ining presence/absence of a diagnosis, we also calculated total
psychiatric symptoms by adding the number of mood, anxiety,
and substance use symptoms participants endorsed (excluding
suicide ideation, so as not to overlap with the suicide ideation
question on the CSS), in order to replicate findings from prior
studies that examined the moderation effect of psychiatric
symptoms.
Data Analytic Plan
Chi-square analyses and independent samples t-tests were used
to assess baseline differences in demographics, mental health
variables, and lifetime exposure to a SA/SD between adoles-
cents who did versus did not make a SA during the follow-up
period. Logistic regression analyses were used to examine the
interaction between exposure to a SA/SD and prior psychiatric
diagnosis in predicting a SA at follow-up. First, lifetime expo-
sure to a SA/SD and prior psychiatric disorders were entered
into the first block of the regression, adjusting for sex, race/
ethnicity, lifetime SA, and prior suicide ideation (Model 1).
Next, the interaction between exposure to a SA/SD and prior
psychiatric disorders was added to the model (Model 2). Addi-
tionally, we conducted a second logistic regression replacing
presence/absence of a prior psychiatric disorder with total psy-
chiatric symptoms. To examine the effects of exposure based
on relationship to the adolescent (i.e., family SA/SD, peers/
friends SA/SD, and others SA/SD) and of cumulative exposure
to SA/SD among vulnerable adolescents, we conducted four
logistic regressions stratified by the presence or absence of a
prior psychiatric disorder, adjusting for sex, race/ethnicity, life-
time SA, and prior suicide ideation. Analyses were conducted
using SPSS, version 25.
Results
At baseline, 78 (15%) participants reported a lifetime SA,
and 172 (33%) reported suicide ideation in the previous
3 months. 130 (25%) Participants met criteria for at least

369Child Psychiatry & Human Development (2022) 53:365–374
1 3
one prior mood, anxiety, or substance use disorder. 40 (8%)
Participants made a SA between baseline and follow-up.
300 (58%) Participants knew at least one person who had
attempted suicide and/or died by suicide in their lifetime
(range 1–19 across participants); 223 (43%) participants
reported exposure to a SA (range 1–15 across participants),
and 156 (30%) were exposed to SDs (range 1–5 across par-
ticipants). Types of exposure by relationship to the adoles-
cent included family SA (N = 34; 7% of total sample), family
SD (N = 25; 5%), peers/friends SA (N = 190; 37%), peers/
friends SD (N = 119; 23%), others SA (N = 21; 4%), and oth-
ers SD (N = 47; 9%).
Of the overall sample, 89 (17%) participants had both
a previous psychiatric diagnosis and also reported lifetime
exposure to a SA/SD. The correlation between having a
previous diagnosis and lifetime exposure to a SA/SD was
statistically significant but low (Cramer’s V = 0.12, p < 0.01).
Baseline Characteristics, Lifetime Exposure toaSA/
SD, andSuicide Attempts atFollow‑up
There were no significant differences by sex or racial/ethnic
group between adolescents who reported a SA at follow-up
and those who did not (see Table1). There were differences
between those who did (vs. did not) report a follow-up SA
in endorsement of prior suicide ideation (65 % vs. 31 %), a
lifetime SA (50% vs. 12%), and prior psychiatric disorders
(53% vs. 23%), respectively, at baseline. Additionally, there
were significant differences in the percentage of adolescents
with a prior mood (30% vs. 11%), anxiety (38% vs. 17%), or
substance use (13% vs. 4%) disorder, respectively, between
those who made a SA since baseline and those who did not.
Three independent samples t-tests found significant differ-
ences in mood (M = 3.92, SD = 4.87 vs. M = 1.31, SD = 3.33),
anxiety (M = 9.25, SD = 9.72 vs. M = 4.99, SD = 6.77), and
total psychiatric symptoms (M = 13.89, SD = 14.79 vs.
M = 6.86, SD = 9.67) between adolescents who reported a
SA vs. those who did not report a SA at follow-up.
Chi-square tests of Independence were used to examine
differences between adolescents that reported a SA at follow-
up vs. no SA at follow-up (see Table1). Adolescents with
and without a lifetime exposure to a SA, a SD or both (SA/
SD) did not differ in their report of SAs during the follow-
up period. However, a significantly higher percentage of
adolescents who attempted suicide at follow-up reported
exposure to a family SA (15%) than did those who did not
attempt suicide at follow-up (6%). There was also a sig-
nificantly higher percentage of adolescents who attempted
suicide at follow-up that reported exposure to a family SA/
SD (25%) than did those who did not attempt at follow-up
(10%). An independent samples t-test found a non-signifi-
cant trend towards a difference in cumulative exposure to
SA/SD between adolescents that reported a SA at follow-
up (M = 1.90, SD = 3.29) vs. those who did not (M = 1.29,
SD = 1.90), p = 0.07.
Interaction Effects Between Lifetime Exposure
toaSA/SD andPsychiatric Disorders
In logistic regression analyses (Model 1, Table2), life-
time exposure to a SA/SD was not associated with having
attempted suicide at follow-up. Race/ethnicity (Black vs.
White) (OR = 2.79, 95% CI = 1.12–6.95), having a prior psy-
chiatric disorder (OR = 2.17, 95% CI = 1.03–4.56), prior sui-
cide ideation (OR = 2.37, 95% CI = 1.08–5.23), and a lifetime
SA (OR = 4.83, 95% CI = 2.24–10.41) were associated with
a SA at follow-up. When the interaction was added to the
model (Model 2, Table2), there was a significant interaction
between the lifetime exposure to a SA/ SD and prior psychi-
atric disorders (OR = 5.78, 95 %CI = 1.05–32.02) (Note: this
interaction was not significant when each type of disorder
was examined separately; results available upon request).
The effect of exposure to a SA/SD was then examined
separately for adolescents with and without prior psychi-
atric disorders (Table3). Stratified analyses indicated that
lifetime exposure to a SA/SD was associated with having
attempted suicide at follow-up among adolescents with prior
Table 2 Logistic regression predicting suicide attempt at follow-up
Bolded values are statistically significant at the p<.05 level
+ Reference groups are White (race/ethnicity) and male (sex)
Model 1: Included race/ethnicity, sex, baseline suicide ideation (pre-
vious 3 months), lifetime SA, prior psychiatric disorders, and lifetime
exposure to a SA/SD
Model 2: Included Model 1 and the interaction between lifetime
exposure to a SA/SD and prior psychiatric disorders
Suicide attempt at follow-up
Model 1 Model 2
OR 95% CI OR 95% CI
Race/ethnicity+
Black
Hispanic
Asian
Other
2.79
2.15
1.14
3.95
1.12–6.95
0.83–5.60
0.28–4.68
0.95–16.42
3.00
2.16
1.19
4.19
1.18–7.58
0.82–5.71
0.28–5.02
1.01–17.46
Sex+1.05 0.48–2.32 1.04 0.47–2.31
Prior suicide ideation 2.37 1.08–5.23 2.19 0.98–4.92
Lifetime SA 4.83 2.24–10.41 5.80 2.60–12.94
Prior psychiatric disorders 2.17 1.03–4.56 0.63 0.14–2.75
Lifetime exposure to a SA/
SD
1.50 0.71–3.22 0.77 0.29–2.03
Interaction
Exposure by psychiatric
disorders
– – 5.78 1.05–32.02

370 Child Psychiatry & Human Development (2022) 53:365–374
1 3
psychiatric disorders (OR = 4.41, 95% CI = 1.04–18.80),
but not among those without prior psychiatric disor-
ders (OR = 0.82, 95% CI = 0.31–2.18). Among adoles-
cents who did not have a prior psychiatric disorder, being
Black (OR = 10.92, 95% CI = 2.12–56.13), Hispanic
(OR = 8.77, 95% CI = 1.67–46.21), or other (OR = 8.99, 95%
CI = 1.15–70.32) significantly predicted a SA at follow-up
compared to White adolescents. However, race/ethnicity did
not predict a SA at follow up among adolescents with a prior
psychiatric disorder.
Additionally, to replicate prior findings, we ran the same
models using total psychiatric symptoms instead of pres-
ence/absence of a prior psychiatric disorder. Exposure to a
SA/SD (OR = 1.46, 95% CI = 0.64–3.33), total psychiatric
symptoms (OR = 1.00, 95% CI = 0.94–1.07), and the inter-
action between exposure to a SA/SD and total psychiatric
symptoms (OR = 1.03, 95% CI = 0.96–1.10) were not sig-
nificant predictors of a SA at follow-up.
Exposure Based ontheRelationship
totheAdolescent
We repeated the stratified analysis presented in Table3 to
examine each exposure based on the relationship to the
adolescent as a predictor of adolescent SAs at follow-up
(Table4). Among adolescents with prior psychiatric disor-
ders, only lifetime exposure to a family SA/SD was associ-
ated with a follow-up SA (OR = 4.14, 95% CI = 1.01–16.99)
This association was not significant among adolescents with-
out prior psychiatric disorders.
Lifetime exposure to peers/friends SA/SD was not asso-
ciated with a SA at follow-up among adolescent with prior
psychiatric disorders (OR = 2.52, 95% CI = 0.77–8.22)
and without prior psychiatric disorders (OR = 0.86, 95%
CI = 0.32–2.32). Similarly, lifetime exposure to others’ SA/
SD was not associated with a SA at follow-up among ado-
lescents with prior psychiatric disorders (OR = 0.84, 95%
CI = 0.14–5.22) and without prior psychiatric disorders
(OR = 0.60, 95% CI = 0.07–4.85).
Table 3 Logistic regressions predicting suicide attempt at follow-up
stratified by the presence/absence of prior psychiatric disorders at
baseline
Bolded values are statistically significant at the p<.05 level
+ Reference groups are White (race/ethnicity) and male (sex)
The model was adjusted for race/ethnicity, sex, baseline suicide idea-
tion (previous 3 months), and lifetime SA
Suicide attempt at follow-up
OR 95% CI
No prior psychiatric disorders
Lifetime exposure to a SA/SD 0.82 0.31–2.18
Race/ethnicity+
Black
Hispanic
Asian
Other
10.92
8.77
3.42
8.99
2.12–56.13
1.67–46.21
0.29–40.54
1.15–70.32
Sex+0.95 0.36–2.52
Prior suicide ideation 1.92 0.63–5.82
Lifetime SA 4.92 1.47–16.50
Prior psychiatric disorders
Lifetime exposure to a SA/SD 4.41 1.04–18.80
Race/ethnicity+
Black
Hispanic
Asian
Other
0.95
0.58
0.52
5.72
0.22–4.15
0.12–2.72
0.07–3.66
0.37–88.87
Sex+1.39 0.30–6.35
Prior suicide ideation 3.11 0.76–12.75
Lifetime SA 8.90 2.58–30.64
Table 4 Logistic regressions predicting suicide attempt at follow-
up stratified by the presence/absence of prior psychiatric disorders at
baseline with lifetime exposure to a family SA/SD as the predictor
Bolded values are statistically significant at the p<.05 level
+ Reference groups are White (race/ethnicity) and male (sex)
The model was adjusted for race/ethnicity, sex, baseline suicide idea-
tion (previous 3 months), and lifetime SA
Suicide attempt at
follow-up
OR 95% CI
No prior psychiatric disorders
Lifetime exposure to a family SA/SD 1.98 0.50–7.85
Race/ethnicity+
Black
Hispanic
Asian
Other
11.19
8.75
3.46
8.22
2.17–57.58
1.66–46.03
0.29–41.02
1.02–66.46
Sex+0.91 0.34–2.44
Prior suicide ideation 1.91 0.63–5.78
Lifetime SA 4.78 1.45–15.71
Prior psychiatric disorders
Lifetime exposure to a family SA/SD 4.14 1.01–16.99
Race/ethnicity+
Black
Hispanic
Asian
Other
0.65
0.29
0.45
4.26
0.14–2.95
0.05–1.68
0.06–3.23
0.31–57.98
Sex+1.52 0.34–6.80
Prior suicide ideation 3.17 0.77–13.04
Lifetime SA 7.02 2.11–23.41

371Child Psychiatry & Human Development (2022) 53:365–374
1 3
Cumulative Exposure toSA/SD
In an analysis stratified by prior psychiatric disorders,
cumulative exposure to SA/SD was not significantly associ-
ated with a SA at follow-up either among adolescents with
prior psychiatric disorders (OR = 1.10, 95% CI = 0.93–1.31)
or among adolescents without prior psychiatric disorders
(OR = 1.06, 95% CI = 0.80–1.41).
Discussion
The present study found that lifetime exposure to a SA/
SD was associated with SA risk among adolescents with
a prior psychiatric disorder, but not among adolescents
without a prior disorder, notably when the person who
attempted suicide or died by suicide was a family mem-
ber. Furthermore, we did not find an association between
cumulative exposure to SAs and SDs and risk of attempt-
ing suicide among adolescents with a psychiatric disorder.
These findings support a theoretical framework in which
adolescents with preexisting psychiatric vulnerability have
a higher risk of attempting suicide when exposed to a SA/
SD than those without psychiatric vulnerability, but this
risk may not increase based on the number of exposures.
One possible explanation is that adolescents with a psy-
chiatric disorder may be more likely to be thinking about
suicide or may already have thought about suicide. As per
the Interpersonal Theory of Suicide, their exposure to sui-
cidal behaviors may have increased their acquired capa-
bility for suicide, prompting the transition from thinking
about suicide to attempting suicide [18]. Being exposed to
suicidal behaviors may contribute to acquiring the capa-
bility for suicide by habituating adolescents to suicidal
behaviors, decreasing their fear of death and subsequently
increasing their risk for attempting suicide. Another pos-
sibility is that knowing someone who has attempted or
died by suicide provides exposure to a potential method
of attempt. The current study did not examine methods
used among people known by the adolescents; however,
the only study that has examined this possibility found a
similarity in suicide methods among adolescents exposed
to the suicide of a parent (albeit with a small sample size
of N = 12) [30]. Future research should consider the simi-
larity between the SA/SD methods used by others and the
methods used by the adolescent to test the imitation or
modeling behavior hypothesis. A third possibility is that
exposed adolescents develop an attentional bias toward
suicide-related stimuli [31], in which they have a height-
ened awareness of and are more likely to focus on informa-
tion relating to suicide when distressed. This attentional
bias could put them at risk for future SAs, as has been
demonstrated among suicide attempters [32], though
inconsistently [33]. Incorporating behavioral paradigms
in addition to self-report measures may provide valuable
information on differences in attentional bias before and
after exposure to a SA or SD.
When we examined the association between exposure
and SA at follow-up using the total number of psychi-
atric symptoms in our regression model, we found that
total psychiatric symptoms, exposure to a SA/SD, and the
interaction were not associated with risk for attempting
suicide. This aligns with previous findings by Nanayakkara
etal. [22] and Swanson and Colman [23], in which sever-
ity of depressive symptoms did not influence the rela-
tion between exposure to suicidal behavior and risk of
SA among adolescents. Our findings indicate that among
adolescents who have been exposed to a lifetime SA/SD,
having a prior psychiatric diagnosis is more meaningful
than symptom severity when evaluating risk for SA. The
absence of an effect of cumulative exposure to SAs and
SDs is congruent with a previous finding that the number
of exposures to SDs or SAs was unrelated to the adoles-
cent’s SA [7]. While inconsistent with Mars etal.’s [21]
finding that exposure to the self-injury of both family and
friend posed greater risk for an adolescent’s SA than being
exposed to the self-injury of either family or friend alone,
our study coded the number of exposures to SA/SD as
a continuous variable and did not collect information on
non-suicidal self-injury, which may account for the differ-
ence in outcome.
Our findings add to mixed evidence on the effects of
exposure to a family vs. peer’s SA/SD by suggesting that
among psychiatrically vulnerable adolescents, exposure to a
family member but not peer’s SA/SD is associated with risk
of a SA. It is unclear how much of the relationship between
exposure to a family member’s suicidal behavior and the
adolescent’s SA is related to closeness of the relationship,
familial environmental stressors, or genetic factors. Future
studies should identify intrapersonal, interpersonal, and
environmental factors that may explain these varied find-
ings. For instance, accounting for shared home environments
and other environmental factors of the family member who
attempted or died by suicide may be one way to assess the
influence of genetic factors on risk for SA [34].
Novel information emerged in our stratified analyses.
Being Black, Hispanic, or Other race/ethnicity significantly
predicted a follow-up SA among adolescents without a
prior psychiatric disorder, but not among adolescents with
a prior disorder. These results suggest that in the absence of
a prior psychiatric disorder, there are racial disparities in the
adolescent’s risk for attempting suicide. These disparities
may be related to unique experiences among Black, His-
panic, and Other adolescents that are not shared with White
adolescents, such as racial/ethnic discrimination, systemic
inequalities, and acculturative stress [35]. These stressful

372 Child Psychiatry & Human Development (2022) 53:365–374
1 3
experiences may make them more likely to attempt suicide
than White adolescents. In fact, the 2019 Youth Risk Behav-
ior Survey (YRBS) of high school students found that Black
students had a higher 12-month prevalence of SAs (11.8%)
than White students (7.9%). While there was no significant
difference between the percentage of Hispanic students and
White students who attempted suicide in the 2019 survey
[36], YRBS data from previous years [37] indicated a higher
prevalence of SAs among Hispanic compared to White stu-
dents. Furthermore, Black and Hispanic youth are at greater
risk of experiencing certain types of traumatic events, such
as childhood maltreatment or low socio-economic status,
and this may also help to explain racial/ethnic differences in
SAs [38, 39]. Our findings are also consistent with research
on youth suicide decedents in which minority youth were
significantly less likely than White youth to have a current
mental health problem or be in treatment [40]. Drawing
conclusions from our data on racial and ethnic differences
in SAs is difficult, as the current study did not adjust for
socioeconomic status, experiences of racial discrimination
or acculturative stress, or other traumatic events that may
explain differential risk for SAs.
Limitations
Study limitations include lack of information about the
temporal sequence of exposure to a SA/SD and SAs at fol-
low-up, given that most adolescents did not report a date of
exposure to the suicidal behavior of someone they know.
Retrospective reports on lifetime exposure may be subject
to recollection bias. In addition, some confidence intervals
in our logistic regressions were wide, due to the relatively
small number of adolescents in the sample who attempted
suicide between baseline and follow-up. Another limitation
is the low prevalence rates of some of our exposure variables
(e.g., exposure to a family SA/SD, exposure to others SA/
SD), which may have biased the results, especially in the
stratified analyses [41]. Predictor variables with higher prev-
alence would likely provide more robust findings. We did not
obtain data on exposure to other stressful life events, such
as non-suicide-related deaths of a close relative or friend,
nor did we assess for baseline post-traumatic stress symp-
toms, which have been shown to be associated with SAs
[42]. Finally, the data were collected over two decades prior
to our analyses. Since then, rates of suicide among adoles-
cents and young adults have steadily increased [2]; and the
growing accessibility and use of social media has expanded
connectedness but can also potentially facilitate communica-
tion about media coverage of suicide deaths, [43], suicidal
thinking [44], and suicidal behavior [45].
Summary
This study highlights the moderating role of prior psychiat-
ric history in the relationship between exposure to a SA or
SD and risk of attempting suicide among adolescents – par-
ticularly when exposure is to a family member’s SA or SD.
Adolescents recruited from high schools in the New York
City metropolitan area as part of a two-stage screening were
assessed for a lifetime SA history and recent suicide idea-
tion and completed a computerized diagnostic interview to
assess for a mood, anxiety, or substance use disorder. They
were followed up 4–6years later and assessed for a SA
since baseline and also for exposure to the SA and/or SD of
someone they knew. Exposure to a SA/SD predicted ado-
lescents’ own SAs at follow-up, but only among those with
a prior psychiatric diagnosis and not among those without
a diagnosis, adjusting for sex, race/ethnicity, lifetime SA
history and recent suicide ideation at baseline. Follow-up
analyses suggested that this was the case for exposure to
the SA/SD of a family member but not of a peer. While
not originally hypothesized, race/ethnicity was a predictor
of a SA at follow-up among adolescents without a baseline
psychiatric diagnosis but not among those with a diagnosis.
Specifically, identifying as Black, Hispanic, or Other race/
ethnicity (compared to White) predicted a SA at follow-up.
A limitation of the study is that exposure to a SA was only
assessed at follow up and not at baseline, making it difficult
to establish temporal order of the relationship. Nevertheless,
these findings suggest that adolescents exposed to a SA or
SD, especially of a family member, may require additional
support if they have a prior history of a psychiatric disorder
to reduce their risk of attempting suicide.
Acknowledgements This research was made possible by the provi-
sion of data by New York State Psychiatric Institute and its research
affiliates (via David Shaffer, retired). This work was funded by Grant
R49/CCR 202598 from the Centers for Disease Control, NIMH Grant
P30 MH 43878, by a Grant from the Carmel Hill Foundation, to David
Shaffer, and by NIH Grant MH091873 (Miranda).
Compliance with Ethical Standards
Conflict of interest The authors declare that they have no conflicts of
interest.
Ethical Approval All procedures performed in studies involving human
participants were in accordance with the Ethical Standards of the
Institutional and/or National Research Committee and with the 1964
Helsinki Declaration and its later amendments or comparable ethical
standards.

373Child Psychiatry & Human Development (2022) 53:365–374
1 3
References
1. Glenn CR, Kleiman EM, Kellerman J, Pollak O, Cha CB, Esposito
EC etal (2020) Annual Research Review: a meta-analytic review
of worldwide suicide rates in adolescents. J Child Psychol Psy-
chiatry 61:294–308
2. Curtin SC, Heron MP (2019) Death rates due to suicide and
homicide among persons aged 10–24: United States, 2000–
2017.NCHS Data Brief 352:1-8
3. Insel BJ, Gould MS (2008) Impact of modeling on adolescent
suicidal behavior. Psychiatr Clin N Am 31:293–316
4. Hawton K, Hill NT, Gould M, John A, Lascelles K, Robinson J
(2019) Clustering of suicides in children and adolescents. Lancet
Child Adolesc Health 4:58–67
5. Nock MK, Green JG, Hwang I, McLaughlin KA, Sampson NA,
Zaslavsky AM, Kessler RC (2013) Prevalence, correlates, and
treatment of lifetime suicidal behavior among adolescents: results
from the National Comorbidity Survey Replication Adolescent
Supplement. JAMA Psychiatry 70:300–310
6. Goldston DB, Daniel SS, Erkanli A, Heilbron N, Doyle O, Weller
B, Sapyta J (2015) Suicide attempts in a longitudinal sample of
adolescents followed through adulthood: evidence of escalation.
J Consult Clin Psychol 83:253–264
7. Burke AK, Galfalvy H, Everett B, Currier D, Zelazny J, Oquendo
MA etal (2010) Effect of exposure to suicidal behavior on suicide
attempt in a high-risk sample of offspring of depressed parents. J
Am Acad Child Adolesc Psychiatry 49:114–121
8. Cerel J, Roberts TA (2005) Suicidal behavior in the family and
adolescent risk behavior. J Adolesc Health 36:352.e8-352.e14
9. Geulayov G, Gunnell D, Holmen TL, Metcalfe C (2012) The
association of parental fatal and non-fatal suicidal behaviour with
offspring suicidal behaviour and depression: a systematic review
and meta-analysis. Psychol Med 42:1567–1580
10. Kuramoto SJ, Runeson B, Stuart EA, Lichtenstein P, Wilcox HC
(2013) Time to hospitalization for suicide attempt by the timing
of parental suicide during offspring early development. JAMA
Psychiatry 70:149–157
11. Mittendorfer-Rutz E, Rasmussen F, Wasserman D (2008) Familial
clustering of suicidal behaviour and psychopathology in young
suicide attempters. Soc Psychiatry Psychiatr Epidemiol 43:28–36
12. Wilcox HC, Arria AM, Caldeira KM, Vincent KB, Pinchevsky
GM, O’Grady KE (2010) Prevalence and predictors of persistent
suicide ideation, plans, and attempts during college. J Affect Dis-
ord 127:287–294
13. Abrutyn S, Mueller AS (2014) Are suicidal behaviors contagious
in adolescence? Using longitudinal data to examine suicide sug-
gestion. Am Sociol Rev 79:211–227
14. Feigelman W, Gorman BS (2008) Assessing the effects of peer
suicide on youth suicide. Suicide Life Threat Behav 38:181–194
15. Gould MS, Lake AM, Kleinman M, Galfalvy H, Chowdhury S,
Madnick A (2018) Exposure to suicide in high schools: impact on
serious suicidal ideation/behavior, depression, maladaptive coping
strategies, and attitudes toward help-seeking. Int J Environ Res
Public Health 15:455
16. Gould MS, Kleinman MH, Lake AM, Forman J, Midle JB (2014)
Newspaper coverage of suicide and initiation of suicide clusters in
teenagers in the USA, 1988–96: a retrospective, population-based,
case–control study. Lancet Psychiatry 1:34–43
17. Bridge JA, Greenhouse JB, Ruch D, Stevens J, Ackerman J, Shef-
tall AH etal (2020) Association between the release of Netflix’s
13 Reasons Why and suicide rates in the United States: an inter-
rupted time series analysis. J Am Acad Child Adolesc Psychiatry
59:236–243
18. Van Orden KA, Witte TK, Cukrowicz KC, Braithwaite SR, Selby
EA, Joiner TE Jr (2010) The interpersonal theory of suicide. Psy-
chol Rev 117:575–600
19. Brent DA, Baugher M, Bridge J, Chen T, Chiappetta L (1999)
Age- and sex-related risk factors for adolescent suicide. J Am
Acad Child Adolesc Psychiatry 38:1497–1505
20. Shaffer D, Gould MS, Fisher P, Trautman P, Moreau D, Kleinman
M, Flory M (1996) Psychiatric diagnosis in child and adolescent
suicide. Arch Gen Psychiatry 53:339–348
21. Mars B, Heron J, Klonsky ED, Moran P, O’Connor RC, Till-
ing K etal (2019) Predictors of future suicide attempt among
adolescents with suicidal thoughts or non-suicidal self-harm: a
population-based birth cohort study. Lancet Psychiatry 6:327–337
22. Nanayakkara S, Misch D, Chang L, Henry D (2013) Depression
and exposure to suicide predict suicide attempt. Depress Anxiety
30:991–996
23. Swanson SA, Colman I (2013) Association between exposure
to suicide and suicidality outcomes in youth. Can Med Assoc J
185:870–877
24. Wong JP, Stewart SM, Ho SY, Rao U, Lam TH (2005) Exposure
to suicide and suicidal behaviors among Hong Kong adoles-
cents. Soc Sci Med 61:591–599
25. Ali MM, Dwyer DS, Rizzo JA (2011) The social contagion
effect of suicidal behavior in adolescents: does it really exist? J
Ment Health Policy Econ 14:3–12
26. Chan S, Denny S, Fleming T, Fortune S, Peiris-John R, Dyson
B (2017) Exposure to suicide behaviour and individual risk of
self-harm: findings from a nationally representative New Zea-
land high school survey. Aust N Z J Psychiatry 52:349–356
27. Shaffer D, Scott M, Wilcox H, Maslow C, Hicks R, Lucas CP
etal (2004) The Columbia Suicide Screen: validity and reli-
ability of a screen for youth suicide and depression. J Am Acad
Child Adolesc Psychiatry 43:71–79
28. Miranda R, Scott M, Hicks R, Wilcox HC, Munfakh JLH,
Shaffer D (2008) Suicide attempt characteristics, diagnoses,
and future attempts: comparing multiple attempters to single
attempters and ideators. J Am Acad Child Adolesc Psychiatry
47:32–40
29. Shaffer D, Fisher P, Dulcan MK, Davies M, Piacentini J,
Schwab-Stone ME etal (1996) The NIMH Diagnostic Inter-
view Schedule for Children Version 2.3 (DISC-2.3): descrip-
tion, acceptability, prevalence rates, and performance in the
MECA study. J Am Acad Child Adolesc Psychiatry 35:865–877
30. Lu TH, Chang WT, Lin JJ, Li CY (2011) Suicide method runs
in families: a birth certificate cohort study of adolescent suicide
in Taiwan. Suicide Life Threat Behav 41:685–690
31. Wenzel A, Beck AT (2008) A cognitive model of suicidal
behavior: theory and treatment. Appl Prev Psychol 12:189–201
32. Cha CB, Najmi S, Park JM, Finn CT, Nock MK (2010) Atten-
tional bias toward suicide-related stimuli predicts suicidal
behavior. J Abnorm Psychol 119:616–622
33. Wilson KM, Millner AJ, Auerbach RP, Glenn CR, Kearns JC,
Kirtley OJ etal (2019) Investigating the psychometric properties
of the Suicide Stroop Task. Psychol Assess 31:1052–1061
34. Brent DA, Melhem N (2008) Familial transmission of suicidal
behavior. Psychiatr Clin N Am 31:157–177
35. Garcia Coll C, Lamberty G, Jenkins R, McAdoo HP, Crnic K,
Wasik BH, Garcia HV (1996) An integrative model for the study
of developmental competencies in minority children. Child Dev
67:1891–1914
36. Underwood JM, Brener N, Thornton J, Harris WA, Bryan LN,
Shanklin SL etal (2020) Youth risk behavior surveillance—
United States, 2019. MMWR Surveill Summ 69:1–114
37. Kann L, McManus T, Harris WA, Shanklin SL, Flint KH, Hawk-
ins J etal (2016) Youth risk behavior surveillance—United
States, 2015. MMWR Surveill Summ 65:1–174

374 Child Psychiatry & Human Development (2022) 53:365–374
1 3
38. Roberts AL, Gilman SE, Breslau J, Breslau N, Koenen KC
(2011) Race/ethnic differences in exposure to traumatic events,
development of post-traumatic stress disorder, and treatment-
seeking for post-traumatic stress disorder in the United States.
Psychol Med 41:71–83
39. Wildeman C, Emanuel N, Leventhal JM, Putnam-Hornstein E,
Waldfogel J, Lee H (2014) The prevalence of confirmed mal-
treatment among US children, 2004 to 2011. JAMA Pediatr
168:706–713
40. Lee CS, Wong YJ (2020) Racial/ethnic and gender differences
in the antecedents of youth suicide. Cult Divers Ethn Minor
Psychol 26:532–543
41. Vittinghoff E, McCulloch CE (2007) Relaxing the rule of ten
events per variable in logistic and Cox regression. Am J Epide-
miol 165:710–718
42. Wilcox HC, Storr CL, Breslau N (2009) Posttraumatic stress
disorder and suicide attempts in a community sample of urban
American young adults. Arch Gen Psychiatry 66:305–311
43. Dunlop SM, More E, Romer D (2011) Where do youth learn about
suicides on the Internet, and what influence does this have on
suicidal ideation? J Child Psychol Psychiatry 52:1073–1080
44. Carlyle KE, Guidry JP, Williams K, Tabaac A, Perrin PB (2018)
Suicide conversations on InstagramTM: contagion or caring? J
Commun Healthc 11:12–18
45. Luxton DD, June JD, Fairall JM (2012) Social media and suicide:
a public health perspective. Am J Public Health 102:S195–S200
Publisher’s note Springer Nature remains neutral with regard to
jurisdictional claims in published maps and institutional affiliations.