Suicide risk among cannabis, heroin and cocaine users

Vol.:(0123456789)
In e na ional Jou nal o Men al Heal h and Addic ion
h ps://doi.o g/10.1007/s11469-025-01460-y
ORIGINAL ARTICLE
Suicide Risk Among Cannabis, He oin andCocaine Use s
MiguelZaca íasPé ezSosa1· Diegode‑la‑Vega‑Sánchez2 · Se gioSanz–Gómez3·
GlennMel in4· AnaSánchez–Gómez5· JulioA.Guija6· Ma iaA.Oquendo7·
LucasGine 3
Accep ed: 19 Feb ua y 2025
© The Au ho (s) 2025
Abs ac
Indi iduals who use subs ances such as cocaine, cannabis, he oin, alcohol and nico ine a e
a inc eased isk o suicidal beha io . Ou objec i e is o examine he ela ionship be ween
subs ance use and suicide. An obse a ional psychological au opsy s udy design was used
o compa e subs ance use in indi iduals who died by suicide and indi iduals who died by
o he manne s. Subjec s we e ec ui ed a he Medical Fo ensic Ins i u e o Se illebe ween
2006 and 2018. The o al sample was 593 deceden s including hose who had died by sui-
cide (n = 412) and hose who died suddenly om non-suicidal causes (n = 181). We de e -
mined he associa ion be ween subs ance (cannabis, he oin, cocaine, alcohol and nico ine)
use and suicide. We also analyzed he beha io al associa ion be ween he use o any one
o hese subs ances (cannabis, he oin, cocaine, alcohol, and nico ine), as well as he use o
mo e han one subs ance a a ime and manne o dea h (suicide s no suicide). We es ab-
lished h ee ca ego ies o subs ance use: ne e -used, li e ime use s (p io use bu no in he
pas mon h), and pas mon h use s (use wi hin he las mon h). This s udy ound ha dece-
den s who died by suicide we e mo e likely o ha e subs ance use compa ed wi h hose
who died by o he means. Use o cocaine, cannabis, o he oin wi hin he las mon h p io
o dea h was associa ed wi h inc eased isk o suicide (X2 = 13.050; d = p = 0.001). O he
subs ances examined, he s onges associa ion was be ween pas mon h cocaine use and
suicide a e adjus ing o Axis I and Axis II diso de s. Ou s udy suppo s he ela ionship
be ween he useo illici subs ances and suicide. The use o subs ances, independen o he
ype, was associa ed wi h a g ea e isk o suicide. Cocaine use had he s onges associa-
ion wi h suicide.
Keywo ds Suicide· Cocaine· Cannabis· He oin· Alcohol· Tobacco· Psychological
au opsy
Globally, app oxima ely 800,000 people die by suicide e e y yea (Suicide, 2022). In
Spain, he suicide a e in 2020 was 8 suicides/100,000 habi an s, compa ed o 10.25 in he
Eu ozone (Minis e io de Sanidad. Gobie no de España, 2024). Acco ding o he Na ional
Ins i u e o S a is ics, suicide con inues o be he leading cause o iolen dea h, wi h 3952
dea hs in 2023, which is g ea e han he 1806 indi iduals who died in mo o ehicle acci-
den s (8.17/100.000 and 3.73/100.000 espec i ely) (INE, 2024). Risk ac o s o suicide
Ex ended au ho in o ma ion a ailable on he las page o he a icle
In e na ional Jou nal o Men al Heal h and Addic ion
include olde age, male sex, low economic s a us, physical illness, and p e ious suicide
a emp s(Chan e al., 2016; G unbaum e al., 2000; Haw on & an Hee ingen, 2009; Haw-
on e al., 2013; Nock e al., 2008). Howe e he e is also a body o esea ch iden i ying
modi iable isk ac o s o suicide, including impulsi eness, agg essi eness, men al heal h
diso de s, suicidal idea ion, and subs ance use. A comp ehensi e unde s anding o he ela-
i e impo ance o hese suicide isk ac o s is he e o e essen ial o e ec i e suicide p e-
en ion s a egies.
In Spain, he p e alence o subs ance use in he pas mon h is as ollows: alcohol 63.5%,
obacco 33.9%, cannabis 10.5%, cocaine 1.4%, he oin 0.1% (Obse a o io Español de
las D ogas y las Adicciones, 2024). When compa ed o neighbou ing coun ies, Spain is
among he op coun ies o li e ime use o bo h cannabis (Spain 43.7%, EU 29.9%) and
cocaine (Spain 13.3%, EU 5.4%) (Eu opeo, 2024).While Spain has lowe suicide a es
compa ed o o he Eu opean coun ies, suicide emains one o he leading causes o unna -
u al dea h in he coun y (in 2023, 3952 suicides we e eco ded, equi alen o a a e o
8.17 pe 100,000 inhabi an s) (INE, 2024). Fu he mo e, al hough i is es ima ed ha he e
was an a e age 20% d op in Eu opean suicide a es om 2011 o 2019, Spain was one o
he Eu opean coun ies whe e he igu e emained s able (Fico e al., 2023). Consequen ly,
he e is a compelling a ionale o a de ailed and igo ous examina ion o his si ua ion.
The ex an li e a u e indica es a clea associa ion be ween he inc eased isk o suicidal
idea ion and suicide and he use o a ious subs ances, including he oin (Bohne e al.,
2017; Chen e al., 2010; Icick e al., 2018; Kazou e al., 2016), cocaine (Abdalla e al.,
2019; A ias e al., 2016; Bohne e al., 2017; Icick e al., 2018), and cannabis (Abdalla
e al., 2019; A ias e al., 2016; Bohne e al., 2017; Gobbi e al., 2019; Halladay e al.,
2019; Lee e al., 2017). The e is obus da a suppo ing he associa ion be ween subs ance
abuse and an inc eased isk o suicide, wi h he excep ion o psychedelics (Hend icks e al.,
2015). Mo eo e , he eco e y p ocess om addic ion and dependence has been ound o
be associa ed wi h a educed isk o suicide (Manhap a e al., 2017). Con e sely, he use o
legal subs ances such as alcohol and nico ine has been linked o an ele a ed isk o suicide,
in addi ion o an inc eased mo ali y a e and pa ien s wi h alcohol abuse ha e an inc eased
isk o suicide, ega dless o como bidi y wi h o he men al diso de s (Flensbo g-Madsen
e al., 2009; Ami i & Behnezhad, 2020). A ecen me a-analysis e ealed ha he isk a io
(RR) o suicide among indi iduals wi h alcohol use, ela i e o non-alcohol use s was
ound o be 1.65 (Ami i & Behnezhad, 2020). Con e sely, a 2021 me a-analysis e ealed
ha smoke s, bo h cu en and pas smoke s in compa ison o non-smoke s, exhibi ed
an ele a ed isk o suicidal beha iou (OR = 1.74; 95% IC 1.54- 1.96) (Eche e ia e al.,
2021).
In addi ion, he numbe o di e en subs ances used could a ec he isk o suicide
a emp s (Lambe & Bonne , 1996; Landheim e al., 2006). In an epidemiological s udy
o illici subs ance use in adolescen s, he Eu opean School Su ey P ojec on Alcohol and
O he D ugs, he isk o a emp ed suicide nea ly doubled wi h e e y addi ional subs ance
aken (Kokke i e al., 2012). In con as , o his obse a ion, ano he s udy e ealed ha
he e was no signi ican dis inc ion in he isk o suicide a emp be ween indi iduals who
use a single subs ance and indi iduals who use mul iple subs ances (Bakken & Vaglum,
2007).
The majo i y o s udies ha e in es iga ed he associa ion be ween illici subs ance use
and suicide a emp s, a he han analyzing he ela ionship be ween illici subs ance use
and suicide. While he e exis some oxicological s udies on suicide, bu hey ha e no
measu ed clinical isk ac o s, such as men al heal h diso de s o p io suicide a emp s.
(Conne e al., 2017; Da ke e al., 2010; Gibbons e al., 2005; Hadidi e al., 2009; O pana
In e na ional Jou nal o Men al Heal h and Addic ion
e al., 2021). Psychological au opsy s udies o e he ad an age o in es iga ing no only he
psychological p o ile o he deceased, bu also enabling he iden i ica ion o subs ance use,
iden i ying illici subs ance use, and can be s eng hened by accessing oxicology esul s.
The objec i e o his s udy is o analyze he isk o suicide associa ed wi h he use o can-
nabis, he oin, cocaine, alcohol and nico ine by employing a psychological au opsy me hod-
ology. We hypo hesize ha cannabis, he oin, cocaine, alcohol and nico ine use wi hin he
las mon h p io o dea h is associa ed wi h suicide.
Me hodology
This s udy cons i u es a componen o an ongoing psychological au opsy p ojec was ini i-
a ed in 2006 by he Ins i u e o Legal Medicine o Se ille. The p ojec in ol es he in i a-
ion o amily membe s o indi iduals who ha e died by suicide o pa icipa e. The p esen
s udy’s sample comp ised 593 subjec s, including 412 consecu i e indi iduals who died by
suicide and 181 non-suicide sudden dea h con ols (na u al and acciden al dea hs) om he
p o ince o Se ille, Spain be ween 2006 and 2018. The o al popula ion o he p o ince o
Se ille in 2024 was 1,967,746, and he suicide a e o ha yea was 8.28 o e e y 100,000
habi an s o Se ille (Ins i u o Nacional de Es adís ica, 2024).
a o ensic in es iga ion and manda o y au opsy. Dea hs occu ing in p ison o unde
police cus ody a e excluded om he s udy. The cases encompassed indi iduals o e he
age o 18yea s esiding in he p o ince o Se ille who had been ce i ied as ha ing com-
mi ed suicide, and whose ela i es had p o ided consen o a subsequen in e iew. The
con ol g oup comp ised indi iduals o e he age o 18yea s esiding in he p o ince o
Se ille who had been ce i ied as ha ing died suddenly o acciden ally ( uling ou suicide),
and whose ela i es had consen ed o a subsequen in e iew.
Psychological au opsies a e a me hod o ob aining in o ma ion abou he subjec h ough
s uc u ed in e iews o amily membe s, iends, and o he indi iduals wi h whom he
subjec had a close ela ionship (Beskow e al., 1990; Haw on e al., 1998). Following a
de ailed explana ion o he s udy, all ela i es o he deceased who wished o pa icipa e
in he in e iew we e in i ed o do so. O he 1,128 suicides ha occu ed du ing he s udy
pe iod, he ela i es o 412 (36.5% o he o al) ag eed o pa icipa e. The in e iews we e
conduc ed 3 o 15mon hs a e he subjec ’s dea h by psychia is s (n = 2) and psycholo-
gis s (n = 2) who had unde gone aining by he p incipal in es iga o (LG). The S uc u ed
Clinical In e iew o DSM-IV Diso de s o Axis I diso de and Axis II diso de s (SCID-
I and SCID-II) we e adminis e ed o assess DSM-IV (Ame ican Psychia ic Associa ion,
1994) diagnoses o subs ance ela ed diso de s. In e iews we e ollowed by an in e disci-
plina y in es iga o mee ing in ol ing (LG, DVS) o each a consensus diagnosis.
The s udy was app o ed by he e hics commi ee o he Uni e si y o Se ille and was
ca ied ou in acco dance wi h he ecommenda ions o he Decla a ion o Helsinki. W i -
en in o med consen was p o ided by all pa icipan s epo ing on he deceased.
Measu es
The SCID-I and SCID-II o DSM-IV s uc u ed in e iews a e diagnos ic measu es o
DSM-IV axis I and II diso de s (Williams e al., 1992). The eliabili y o bo h in e iews
has been demons a ed in s udies employing psychological au opsy (Kelly & Mann, 1996)
and he ha e been p e iously u ilized by ou g oup (Gine e al., 2013).
In e na ional Jou nal o Men al Heal h and Addic ion
In he p esen s udy, an analysis was conduc ed on he in luence o he use o illegal
subs ances (including cannabis, he oin and cocaine) and legal subs ances (including alco-
hol and nico ine) using he subs ance abuse ca ego y o he SCID-I. This ca ego y is used
o make a diagnosis o abuse in he las wel e mon hs and can speci y i s p esence in
he las mon h. Th ee ca ego ies o use eme ged om he SCID-I: no subs ance abuse in
he las 12mon hs, subs ance abuse in he las 12mon hs, and subs ance abuse in he las
mon h. The applica ion o hese ca ego ies enabled he analysis o a g adien o suicide isk
acco ding o he pa e n o use, and o asce ain whe he he lowes isk was o hose who
had no used in he las 12mon hs, he in e media e isk was o hose who had used in he
las 12mon hs, and he highes isk was o hose who had used in he las mon h.
Fu he mo e, a bespoke ques ionnai e was de eloped in o de o analyse each o he
h ee subs ances in h ee dis inc ca ego ies: ne e use s ( hose who ha e ne e used he
subs ance), li e ime use s ( hose who ha e used he subs ance in he pas bu no in he las
mon h) and las mon h use s ( hose who ha e used he subs ance in he p e ious mon h).
This ques ionnai e p o ided subs ance-speci ic esul s o e a mo e ex ended pe iod o ime
(li e ime).
S a is ical Analysis
The s a is ical analysis was conduc ed using SPSS V26.0. Desc ip i e s a is ics we e es i-
ma ed wi h mean o con inuous a iables and pe cen ages o ca ego ical a iables. Pea -
son Chi-squa e (and Fishe ´s es i Pea son’s Chi-squa ed es was no applicable) and
- es we e used o compa e ca ego ical and con inuous a iables, espec i ely. The odds
a ios (OR) and 95% con idence in e als we e calcula ed o associa ions be ween suicide
and ca ego ies o subs ance use (ne e -use s, li e ime use s, pas mon h use s). To in es i-
ga e he possibili y o a dose– esponse ela ionship be ween subs ance use and suicide, we
employed he Man el–Haenszel (M2) end es using he ¡TCOR mac o. Las ly, a logis ic
eg ession model o con ol o o he a iables ela ed o suicide such as he p esence o
men al diso de s (Axis I and Axis II diso de s) o he use o he mos common subs ances
such as alcohol and obacco, examined he ela ionship be ween subs ance use (independ-
en a iables) and he dependen a iable being manne o dea h (suicide o con ol).
Model 1 included I Axis I and Axis II, while Model II included Axis I, Axis II, alcohol and
nico ine in pas mon h.
Resul s
The e we e no signi ican sociodemog aphic di e ences be ween suicide and con-
ol g oups (See Table 1) excep ha hose in he con ol g oup we e mo e likely o be
employed (suicides 14.6%, con ols 23.8% χ2 = 6.432 p = 0.040).
Those who died by suicide we e mo e likely han con ols o ha e psychia ic ollow-up
(23.2% Vs 12%, χ2 = 9.355, p = 0.025), o ha e a men al heal h diso de (Table2), includ-
ing bo h Axis I (85.7% Vs. 60.8%, χ2 = 45.574 p < 0.001) and Axis II diso de s (64.1% Vs.
38.9%, χ2 = 32.328 p < 0.001). Simila ly, suicide deceden s we e mo e likely o ha e sub-
s ance abuse o illici subs ances (13.1% Vs. 3.3%, χ2 = 13.258, p < 0.001) o dependence
(8.7% Vs. 1.1%, χ2 = 12.216 p < 0.001) diso de in he p e ious mon h, and alcohol abuse
(21.1% Vs. 13.9%, χ2 = 4.266, p = 0.039) o dependence (13.6% Vs. 7.7%, χ2 = 6.333,
p = 0.042) in he p e ious mon h.
In e na ional Jou nal o Men al Heal h and Addic ion
Rega ding subs ance use (ne e -use s, li e ime use s, pas mon h use s), deceden s who
died by suicide and did o did no use subs ances in he pas mon h di e ed on se e al a i-
ables. Compa ed wi h he ne e -use s and li e ime use s, suicide pas mon h use s we e:
younge (mean age 57.03, 61.413, 45.781, F = 33.274 p < 0.001), mo e likely o be men
Table 1 Sociodemog aphic a iables in suicide and con ol g oups
S udy g oups
Suicides
N = 412
(69.48%)
Con ol
N = 181
(30.52%)
To al
593
(100.0%)
d S a is ics
Age Mean 53.769 56.762 54.485 591 = −1.738 p = 0.083
Sex Male 308
(74.8%)
127
(70.2%)
364
(74.1%)
1χ2 = 1.356 p = 0.244
Female 104
(25.2%)
54
(29.8%)
127
(25.9%)
Has a pa ne Has a pa ne 196 (47.6%) 87 (48.1%) 235
(47.9%)
1χ2 = 0.012 p = 0.912
O sp ing Yes 280
(68.3%)
133
(74.7%)
349
(71.7%)
1χ2 = 2.452 p = 0.117
Li ing a ange-
men s
Alone 74 (18.5%) 29 (16.5%) 82
(17.2%)
1χ2 = 0.326 p = 0.568
Employmen s a us Employed 52 (14.6%) 36 (23.8%) 71
(17.0%)
2χ2 = 6.432 p = 0.040
No employed wi h
income
229 (64.1%) 85 (56.3%) 256
(61.4%)
No employed
wi hou income
76 (21.3%) 30
(19.9%)
90
(21.6%)
Table 2 Diagnos ic cha ac e is ics o suicides and con ols
* (Cocaine, cannabis, he oin)
Suicides
N = 412
Con ols
N = 181
χ2, p
Men al heal h ollow-up 82 (23.2%) 14 (12%) 9.355, p = 0.025
Axis I 353 (85.7%) 110 (60.8%) 45.574, p < 0.001
Axis II 261 (64.1%) 70 (38.9%) 32.328, p < 0.001
Li e ime subs ance* abuse 63 (15.3%) 13 (7.2%) 7.400, p = 0.007
Pas mon h subs ance* abuse 54 (13.1%) 6 (3.3%) 13.258, p < 0.001
Li e ime subs ance* dependence 39 (9.5%) 4 (2.2%) 9.845, p = 0.002
Pas mon h subs ance* dependence 36 (8.7%) 2 (1.1%) 12.216, p < 0.001
Li e ime alcohol abuse 110 (26.7%) 38 (21.0%) 2.185, p = 0.139
Pas mon h alcohol abuse 87 (21.1%) 25 (13.9%) 4.266, p = 0.039
Li e ime alcohol dependence 72 (17.5%) 21 (11.6%) 3.281, p = 0.070
Pas mon h alcohol dependence 56 (13.6%) 14 (7.7%) 6.333, p = 0.042
Li e ime obacco consump ion 293 (71.6%) 116 (64.8%) 2.746, p = 0.098
Pas mon h obacco consump ion 214 (57.8%) 88 (50.9%) 2.231, p = 0.128

In e na ional Jou nal o Men al Heal h and Addic ion
(39.4%, 72.5%, 89.6%, χ2 = 65.687 p < 0.001); had ewe child en (2.045, 2.213, 1.591,
F = 5.374 p = 0.005); and we e mo e likely o hold ac i e labo jobs (12.1%, 4.8%, 25.2%,
χ2 = 53.395 p < 0.001).
Subs ance Abuse
Compa ed o con ols, subjec s who died by suicide we e mo e likely o ha e been using a
leas one subs ance (o he han alcohol o nico ine) in he mon h p io o dea h (Z = 3.378.
(p < 0.001) – see Table 3. Compa ed wi h hose who ne e used subs ances, li e ime
use s we e a no g ea e isk o suicide (OR = 0.832, 95% CI = 0.322–2.153), bu pas
mon h use s had a nea ly ou - old isk o suicide (OR = 4.370, 95% CI = 1.843–10.361).
The isk o suicide associa ed wi h pas mon h use o any illici subs ance was p esen
e en a e adjus ing o he p esence o Axis I o II diso de s, model I, (OR = 1.446, IC
95% = 1.934–4.588) and he p esence o Axis I, Axis II, alcohol and nico ine in pas mon h,
model II, (OR = 1.49, IC 95% = 1.006–2.22).
Pas Mon h Use o Any Illici Subs ance (Cannabis, He oin o Cocaine)
Los da os ob enidos con es a p opues a son muy simila es a los ob enidos empelando la
SCID-I. Compa ed o con ols, subjec s who died by suicide we e mo e likely o ha e been
using a leas one subs ance (o he han alcohol o nico ine) in he mon h p io o dea h
(Z = 3.602, p < 0.001) – see Table3. Compa ed wi h hose who ne e used subs ances, li e-
ime use s we e a no g ea e isk o suicide (OR = 1.762, 95% CI = 0.863–3.598), bu pas
mon h use s had a nea ly h ee- old isk o suicide (OR = 2.985, 95% CI = 1.553–7.739).
The isk o suicide associa ed wi h pas mon h use o any illici subs ance) was p esen
e en a e adjus ing o he p esence o Axis I o II diso de s, model I, (OR = 2.010, IC
95% = 1.021–3.959) and he p esence o Axis I, Axis II, alcohol and nico ine in pas mon h,
model II, (OR = 2.266, IC 95% = 1.103–4.654).
Polysubs ance Use (Cannabis and/o He oin and/o Cocaine)
Pas mon h useo mo e han one subs ance (o he han alcohol o nico ine) in he mon h
p io o dea h was only p esen in 27 suicides (6.55%) and 4 con ols (2.21%). The e
was no di e ence in likelihood o polysubs ance use in he pas mon h be ween g oups
(OR = 2.116, 95% CI = 0.704–6.361).
Cannabis
Rela i e o con ols, suicide deceden s we e mo e likely o ha e used cannabis in he pas
mon h (Z = 2.696, p = 0.007). Pas use s compa ed o ne e use s we e a no g ea e isk
o suicide (OR = 1.821, IC 95% = 0.868–3.819), bu pas mon h use s we e a g ea e
isk (OR = 2.276, IC 95% = 1.138–4.554). Pas mon h use o cannabis was associa ed
wi h a g ea e isk o suicide compa ed o ne e use s and pas cannabis use s al hough
a e adjus ing o o he a iables in model I (1.396, 95% CI = 0.694–2.807) and model II
(OR = 1.493, 95% CI = 0.684–3.027) he di e ence ceased o be signi ican .
In e na ional Jou nal o Men al Heal h and Addic ion
Table 3 Subs ance use in suicides and con ols adjus ed o p esence o Axis I and II diso de s
d Pea son’s chi-
squa ed es *
Lineal en-
dency es
(M2)
Adjus ed model
I**
OR (95% C.I.)
Adjus ed model
II***
OR (95% C.I.)
Suicide Con ol OR
Subs ance abuse
(SCID-I)
No subs ance
abuse
(12mon hs)
346 168 1 χ2 = 13.408
(P < 0.001)
Z = 3.378
(p < 0.001)
1.446 (1.934–
4.588)
1.495 (1.006–2.22)
Subs ance abuse
(in he las
12mon hs)
15 7 1.04 (0.416–2.6)
Subs ance abuse in
he las mon h
51 6 4.127 (1.736–
9,809)
Any subs ance
(Cannabis
o He oin o
Cocaine)
Ne e -Use 235 (71.4%) 138 (85.7%) 1 1 χ2 = 13.050
(p = 0.001)
Z = 3.602
(p < 0.001)
2.010 (1.021–
3.959)
2.266 (1.103–4.654)
Li e ime use s 33 (10.0%) 11 (6.8%) 1.762 (0.863–
3.598)
Pas mon h use s 61 (18.5%) 12 (7.5%) 2.9851 (1.553–
5.739)
Cannabis Ne e -Use 239 (75.9%) 136 (86.6%) 1 2 χ2 = 7.573
(p = 0.023)
Z = 2.696
(p = 0.007)
1.396 (0.694–
2.807)
1.493 (0.684–3.027)
Li e ime use s 32 (10.2%) 10 (6.4%) 1.821 (0.868–
3.819)
Pas mon h use s 44 (14.0%) 11(7.0%) 2.276 (1.138–
4.554)
He oin Ne e -Use 282 (94.0%) 148 (98.0%) 1 2 p = 3.576
(p = 0.165)
Z = 1.925
(p = 0.054)
2.411 (0.513–
11.329)
2.640 (0.551–
12.652)
Li e ime use s 4 (1.3%) 1 (0.7%) 2.099 (0.233–
18.952)
Pas mon h use s 14 (4.7%) 2 (1.3%) 3.673 (0.824–
16.381)
Cocaine Ne e -Use 259 (84.9%) 147 (96.1%) 1 2 p = 13.624
(p = 0.001)
Z = 3.489
(p < 0.001)
3.243 (1.096–
9.595)
3.140 (1.041–9.475)
Li e ime use s 12 (3.9%) 2 (1.3%) 3.405 (0.752–
15.424)
Pas mon h use s 34 (11.1%) 4 (2.6%) 4.824 (1.679–
13.864)
In e na ional Jou nal o Men al Heal h and Addic ion
Table 3 (con inued)
d Pea son’s chi-
squa ed es *
Lineal en-
dency es
(M2)
Adjus ed model
I**
OR (95% C.I.)
Adjus ed model
II***
OR (95% C.I.)
Suicide Con ol OR
Alcohol Ne e -Use 88
(22.7%)
41
(23.6%)
1 2 χ2 = 0.146
(p = 0.929)
Z = 0.309
(0.757)
1.000 (0.666–
1.502)
Li e ime use s 35
(9.0%)
17
(9.8%)
0.959
(0.482–1.908)
Pas mon h use s 264
(68.2%)
116
(66.7%)
1.060 (0.690–
1.630)
Nico ine Ne e -Use 113
(30.5%)
58
(33.5%)
12χ2 = 2.816
(p = 0.245)
Z = 1.1979
(p = 0.231)
1.024 (0.692–
1.515)
Li e ime use s 43
(11.6%)
27
(15.6%)
0.817 (0.459–
1.454)
Pas mon h use s 214
(57.8%)
88
(50.9%)
1.248 (0.835–
1.866)
# OR Odds a io; ^CI Con idence in e al, *Fishe ´s es i Pea son’s chi-squa ed Tes no applicable **Model I: Axis I and Axis II, ***Model II: Axis I, Axis II, alcohol and
nico ine in pas mon h
In e na ional Jou nal o Men al Heal h and Addic ion
He oin
Those who died by suicide ended o ha e used he oin in he pas mon h (Z = 1.925,
p = 0.054) compa ed o con ols. Risk o suicide in ne e use s did no di e om isk
in pas he oin use s (OR = 2.099, 95% CI = 0.233–18.952). The pas mon h use o he oin
was no associa ed wi h a g ea e isk o suicide (OR = 3.673, IC 95% = 0.824–16.381)
compa ed o ne e use s and pas he oin use s, which did no change when adjus ing wi h
model I (OR = 2.411, IC 95% = 0.513–11.329) o II (OR = 2.640, 95% IC = 0.551–12.652).
Cocaine
Pas mon h cocaine use was mo e common in suicides ela i e o con ols (Z = 3.489,
p < 0.001). Pas cocaine use s we e no mo e likely o die by suicide ela i e o ne e
use s (OR = 3.405, IC 95% = 0.752–15.424). Howe e , hose who used cocaine du -
ing he mon h be o e dea h we e mo e likely o ha e died by suicide (OR = 4.824, IC
95% = 1.679–13.864). Pas mon h use o cocaine was associa ed wi h g ea e isk o sui-
cide, compa ed o ne e use s and pas cocaine use s which pe sis ed a e adjus ing wi h
model I (OR = 3.243, IC 95% = 1.096–9.595) o II (OR = 3.140, 95% IC = 1.041–9.475).
Alcohol
Pas mon h alcohol use was no mo e common in suicide dea hs ela i e o con ols
(Z = 0.309, p < 0.757). Pas alcohol use s we e no mo e likely o die by suicide ela i e o
ne e use s (OR = 0.959, CI 95% = 0.482–1.908). Pas mon h use o alcohol was no associ-
a ed wi h g ea e isk o suicide (OR = 1.060, CI 95% = 0.690–1.630) compa ed o ne e
use s and pas alcohol use s, which did no change when adjus ing o Axis 1 and 2 diso -
de s (model I) (OR = 1.000, CI 95% = 0.666–1.502).
Nico ine
Pas mon h nico ine use was no mo e common in suicides ela i e o con ols (Z = 1.1979,
p < 0.231). Pas nico ine use s we e no mo e likely o die by suicide ela i e o ne e use s
(OR = 0.817, CI 95% = 0.459–1.454). Pas mon h use o nico ine was no associa ed wi h
a g ea e isk o suicide (OR = 1.248, CI 95% = 0.835–1.866) compa ed o ne e use s
and pas nico ine use s, which did no change when adjus ing o Axis 1 and 2 (model I)
(OR = 1.024, CI 95% = 0.692–1.515).
Discussion
The indings o he p esen s udy sugges a ela ionship be ween he use o cocaine and
cannabis, and suicide. Subjec s who had used cocaine and cannabis in he las mon h p io
o dea h we e ound be a a g ea e isk o suicide han hose who had no used in he pas
mon h o had ne e used.
The s onges associa ion was ound o be be ween pas mon h cocaine use and suicide
wi h he associa ion pe sis ing when adjus ing o Axis I and II diso de s. Mo eo e , pas
In e na ional Jou nal o Men al Heal h and Addic ion
Kelly, J., G eene, M., & Be gman, B. (2018). Beyond abs inence: Changes in indices o quali y o li e
wi h ime in eco e y in a na ionally ep esen a i e sample o U.S. adul s. Alcoholism, Clinical and
Expe imen al Resea ch, 42(4), 770–780. h ps:// doi. o g/ 10. 1111/ ACER. 13604
Kelly, T. M., & Mann, J. J. (1996). Validi y o DSM-III-R diagnosis by psychological au opsy: A com-
pa ison wi h clinician an e-mo em diagnosis. Ac a Psychia ica Scandina ica, 94(5), 337–343.
h ps:// doi. o g/ 10. 1111/J. 1600- 0447. 1996. TB098 69.X
Kokke i, A., e al. (2012). Mul iple subs ance use and sel - epo ed suicide a emp s by adolescen s in
16 Eu opean coun ies. Eu opean Child & Adolescen Psychia y, 21(8), 443–450. h ps:// doi. o g/
10. 1007/ s00787- 012- 0276-7
Lambe , M. T., & Bonne , J. (1996). Cha ac e is ics and six-mon h ou come o pa ien s who use suicide
h ea s o seek hospi al admission. Psychia ic Se ices (Washing on, D. C.), 47(1075–2730), 871–873.
Landheim, A. S., Bakken, K., & Vaglum, P. (2006). Wha cha ac e izes subs ance abuse s who commi
suicide a emp s? Fac o s ela ed o Axis I diso de s and pa e ns o subs ance use diso de s. A
s udy o ea men -seeking subs ance abuse s in No way. Eu opean Addic ion Resea ch, 12(1022-
6877 (P in )), 102–108.
Lannoy, S., e al. (2022). Risk o non- a al suicide a emp in indi iduals wi h subs ance use diso de : The
oles o agg ega e gene ic liabili y and en i onmen al exposu es in a Swedish popula ion-based coho .
Addic ion (Abingdon, England), 117(11), 2943–2952. h ps:// doi. o g/ 10. 1111/ ADD. 15965
Lee, K. H., e al. (2017). Men al heal h, subs ance abuse, and suicide among homeless adul s. Jou nal o
E idence-In o med Social Wo k, 14(4), 229–242. h ps:// doi. o g/ 10. 1080/ 23761 407. 2017. 13162 21
Le -Ran, S., e al. (2014). The associa ion be ween cannabis use and dep ession: A sys ema ic e iew and
me a-analysis o longi udinal s udies. Psychological Medicine, 44(4), 797–810. h ps:// doi. o g/ 10.
1017/ S0033 29171 30014 38
Luoma, J. B., Ma in, C. E., & Pea son, J. L. (2002). Con ac wi h men al heal h and p ima y ca e p o ide s
be o e suicide: A e iew o he e idence. The Ame ican Jou nal o Psychia y, 159(6), 909. h ps:// doi.
o g/ 10. 1176/ APPI. AJP. 159.6. 909
Manhap a, A., S e ano ics, E., & Rosenheck, R. (2017). T ea men ou comes o e e ans wi h PTSD and
subs ance use: Impac o speci ic subs ances and achie emen o abs inence. Physiology & Beha io ,
176(1), 139–148. h ps:// doi. o g/ 10. 1016/j. physb eh. 2017. 03. 040
Man ique-Ga cia, E., e al. (2016). Cannabis, psychosis, and mo ali y: A coho s udy o 50,373 Swedish
men. The Ame ican Jou nal o Psychia y, 173(8), 790–798. h ps:// doi. o g/ 10. 1176/ APPI. AJP. 2016.
14050 637
Ma he s, B., e al. (2013). Mo ali y among people who injec d ugs: A sys ema ic e iew and me a-analysis.
Bulle in o he Wo ld Heal h O ganiza ion, 91(2), 102–123. h ps:// doi. o g/ 10. 2471/ BLT. 12. 108282
Miguel, A., e al. (2019). Sho and long- e m imp o emen s in psychia ic symp oma ology o alida e clin-
ically meaning ul ea men ou comes o cocaine use diso de s. D ug and Alcohol Dependence, 198,
126–132. h ps:// doi. o g/ 10. 1016/J. DRUGA LCDEP. 2019. 01. 046
Minis e io de Sanidad. Gobie no de España. (2024). In o me Mensual Mayo 2024 Línea 024.
Nock, M. K., e al. (2008). Suicide and suicidal beha io . Epidemiologic Re iews, 30(1), 133–154. h ps://
doi. o g/ 10. 1093/ EPIREV/ MXN002
Obse a o io Español de las D ogas y las Adicciones. (2019). Es adís icas 2019: Alcohol, abaco y d ogas
ilegales en España.
Obse a o io Español de las D ogas y las Adicciones. (2024). Obse a o io Español de las D ogas y las
Adicciones. Es adís icas 2024. Alcohol, abaco y d ogas ilegales en Es- paña. Mad id: Minis e io de
Sanidad. Delegación del Gobie no pa a el Plan Nacional sob e D ogas; 2024. 301 p.
O pana, H., e al. (2021). Alcohol and o he d ugs in suicide in Canada: Oppo uni ies o suppo p e en ion
h ough enhanced moni o ing. Inju y P e en ion: Jou nal o he In e na ional Socie y o Child and
Adolescen Inju y P e en ion, 27(2), 194–200. h ps:// doi. o g/ 10. 1136/ INJUR YPREV- 2019- 043504
Pai a, C., e al. (2017). Dep ession, anxie y, hopelessness and quali y o li e in use s o cocaine/c ack in
ou pa ien ea men . T ends in Psychia y and Psycho he apy, 39(1), 34–42. h ps:// doi. o g/ 10. 1590/
2237- 6089- 2015- 0065
Pa a in, R., & Fio i i, A. (2018). Mo ali y ends among cocaine use s ea ed be ween 1989 and 2013 in
No he n I aly: Resul s o a longi udinal s udy. Jou nal o Psychoac i e D ugs, 50(1), 72–80. h ps://
doi. o g/ 10. 1080/ 02791 072. 2017. 13659 76
Peacock, A., e al. (2021). All-cause and cause-speci ic mo ali y among people wi h egula o p oblema ic
cocaine use: A sys ema ic e iew and me a-analysis. Addic ion (Abingdon, England), 116(4), 725–742.
h ps:// doi. o g/ 10. 1111/ ADD. 15239
Plana-Ripoll, O., e al. (2019). A comp ehensi e analysis o mo ali y- ela ed heal h me ics associa ed wi h
men al diso de s: A na ionwide, egis e -based coho s udy. Lance (London, England), 394(10211),
1827–1835. h ps:// doi. o g/ 10. 1016/ S0140- 6736(19) 32316-5

In e na ional Jou nal o Men al Heal h and Addic ion
P es on, K. L., & Eps ein, D. H. (2011). S ess in he daily li es o cocaine and he oin use s: Rela ionship o
mood, c a ing, elapse igge s, and cocaine use. Psychopha macology (Be l), 218(1), 29–37. h ps://
doi. o g/ 10. 1007/ s00213- 011- 2183-x
P ice, C., e al. (2009). Cannabis and suicide: Longi udinal s udy. B i ish Jou nal o Psychia y, 195(6),
492–497. h ps:// doi. o g/ 10. 1192/ bjp. bp. 109. 065227
Rylande , M., Valdez, C., & Nussbaum, A. M. (2014). Does he legaliza ion o medical ma ijuana inc ease
comple ed suicide? Ame ican Jou nal o D ug and Alcohol Abuse, 40(4), 269–273. h ps:// doi. o g/ 10.
3109/ 00952 990. 2014. 910520
S ene-La sen, K., & Rene lo , A. (2017). Con ac wi h p ima y and men al heal h ca e p io o suicide:
A sys ema ic e iew o he li e a u e om 2000 o 2017. 47(1), 9–17. h ps:// doi. o g/ 10. 1177/ 14034
94817 746274
Suicide. (2022). A ailable a : h ps:// www. who. in / news- oom/ ac - shee s/ de ail/ suici de. Accessed 26 No
2022.
Van O den, K. A., e al. (2010). The in e pe sonal heo y o suicide. Psychological Re iew, 117(2), 575–
600. h ps:// doi. o g/ 10. 1037/ a0018 697
Williams, J. B. W., e al. (1992). The s uc u ed clinical in e iew o DSM-III-R (SCID). II. Mul isi e es -
e es eliabili y. A chi es o Gene al Psychia y, 49(8), 630–636. h ps:// doi. o g/ 10. 1001/ ARCHP
SYC. 1992. 01820 08003 8006
Wyco , A., Me ik, J., & T ull, T. (2018). A ec and cannabis use in daily li e: A e iew and ecommen-
da ions o u u e esea ch. D ug and Alcohol Dependence, 191, 223–233. h ps:// doi. o g/ 10. 1016/J.
DRUGA LCDEP. 2018. 07. 001
Publishe ’s No e Sp inge Na u e emains neu al wi h ega d o ju isdic ional claims in published maps and
ins i u ional a ilia ions.
Au ho s and A ilia ions
MiguelZaca íasPé ezSosa1· Diegode‑la‑Vega‑Sánchez2 · Se gioSanz–Gómez3·
GlennMel in4· AnaSánchez–Gómez5· JulioA.Guija6· Ma iaA.Oquendo7·
LucasGine 3
* Diego de-la-Vega-Sánchez
d [email p o ec ed]
1 Vi gen Maca ena Uni e si y Hospi al, Se ille, Spain
2 Depa men o Psychia y, School o Medicine, Uni e sidad de Se illa, Vi gen Maca ena
Uni e si y Hospi al, Se ille, Spain
3 Depa men o Psychia y, School o Medicine, Uni e sidad de Se illa, Se ille, Spain
4 School o Psychology, Facul y o Heal h, Deakin Uni e si y, Bu wood, Aus alia
5 Ins i u e o Legal Medicine o Se illa, Se ille, Spain
6 Ins i u e o Legal Medicine o Se illa, Depa men o Psychia y School o Medicine, Uni e sidad
de Se illa, Se ille, Spain
7 Depa men o Psychia y in hePe elman School o Medicine a  heUni e si y o Pennsyl ania,
Philadelphia, PA, USA