Acute alcohol use and suicide deaths: An analysis of New Zealand coronial data from 2007–2020

Abstract

Aims: Acute alcohol use is a proximal risk factor for suicide. However, the proportion of suicide deaths involving acute alcohol use has
not been quantified in New Zealand. We sought to quantify and characterise the association between acute alcohol use and suicide.
methods: Data for all suicides (≥15 years) between July 2007 and December 2020 were drawn from the National Coronial Information
System. Acute alcohol use was defined as blood alcohol concentration (BAC) >50mg/100mL. Logistic regression was used to compare characteristics between suicide deaths with and without acute alcohol use.
Results: Twenty-six point six percent of suicide deaths involved acute alcohol use. No difference in the association was found by sex (male AOR: 0.87 (95%CI: 0.74,1.02)). Ethnicity differences were identified (Māori AOR: 1.20 (95%CI: 1.01,1.42), Pacific AOR: 1.46 (95%CI: 1.10,2.00)). Those aged 15–54 years had similar risks of suicide involving acute alcohol use, with a lower association in older
age groups.
Conclusions: Acute alcohol use was identified in approximately one quarter of suicides, with stronger associations in those of Māori and Pasifika ethnicity, and those aged <55 years. Acute alcohol use is a significant but modifiable risk factor for suicide in New Zealand.

Full text

New Zealand Medical Journal
Te ara tika o te hauora hapori
2022 Jul 15; 135(1558). ISSN 1175-8716
www.nzma.org.nz/journal ©PMA
article 65
Acute alcohol use and suicide deaths:
an analysis of New Zealand coronial
data from 2007–2020
Rose Crossin, Lana Cleland, Annette Beautrais, Katrina Witt, Joseph M Boden

: Acute alcohol use is a proximal risk factor for suicide. However, the proportion of suicide deaths involving acute alcohol use has
not been quantified in New Zealand. We sought to quantify and characterise the association between acute alcohol use and suicide.
: Data for all suicides (≥15 years) between July 2007 and December 2020 were drawn from the National Coronial Information
System. Acute alcohol use was defined as blood alcohol concentration (BAC) >50mg/100mL. Logistic regression was used to compare
characteristics between suicide deaths with and without acute alcohol use.
: Twenty-six point six percent of suicide deaths involved acute alcohol use. No dierence in the association was found by
sex (male AOR: 0.87 (95%CI: 0.74,1.02)). Ethnicity dierences were identified (Māori AOR: 1.20 (95%CI: 1.01,1.42), Pacific AOR: 1.46
(95%CI: 1.10,2.00)). Those aged 15–54 years had similar risks of suicide involving acute alcohol use, with a lower association in older
age groups.
: Acute alcohol use was identified in approximately one quarter of suicides, with stronger associations in those of Māori
and Pasifika ethnicity, and those aged <55 years. Acute alcohol use is a significant but modifiable risk factor for suicide in New Zealand.
Acute alcohol use is a known proximal risk
factor for suicide,1 and has been shown
to signicantly increase risk of suicide
attempt, particularly at high levels of acute con-
sumption.2 For suicide deaths, reviews nd the
prevalence of acute alcohol use range from 10%
to 69%, differing by population demographics
including age and sex.3 Alcohol may neurocog-
nitively trigger suicide attempts by increasing
impulsivity and disinhibition,4 weakening psy-
chological barriers to suicide attempts,5 or by
increasing despair, and cognitively impairing
efforts to mitigate despair.6 Acute use of alco-
hol is associated with use of more lethal suicide
means,4,5 and may potentiate the effects of other
drugs consumed in overdose,3 thereby reducing
the likelihood of surviving an attempt. These nd-
ings suggest that acute alcohol use should be a
focus for suicide prevention.
Recent coronial data studies in Australia and
South Korea provide data about characteristics
of acute alcohol use in suicide.7–9 In Australia,
around one quarter (26.7%) of suicide decedents
had a blood alcohol concentration (BAC) of
≥0.05g/100mL (the legal drink-driving limit in
Australia); alcohol use prior to suicide was associ-
ated with male sex and use of more lethal means.7
In South Korea, a study of 683 suicide decedents
found that almost one third (28.7%) had a BAC
≥0.08g/100mL (the legal drink-driving limit in
that country, and dened as “intoxication”); acute
alcohol intoxication was associated with having
no underlying medical or psychiatric diagnosis.8
This negative association between intoxication
and psychiatric history has also been identied
in Australia: alcohol use prior to suicide was
associated with acute stress (e.g., relationship
breakdown), but not with psychiatric illness.9
These studies suggest that acute alcohol use may
increase impulsive suicide risk in those without
psychiatric risk factors but who are exposed to an
acute stressor.
Suicide is a signicant public health issue in
New Zealand. The suicide rate in June 2021 was
11.6/100,000 population.10 This rate is higher for
Māori (15.8/100,000), and for those aged 15–24
(11.4 and 22.2 per 100,000 for females and males,
respectively).10 The World Health Organization
emphasises that almost one in ve of all suicides
can be attributed to alcohol use,11 highlighting
alcohol policy as a point of intervention for reduc-
ing suicide.12 Not targeting alcohol represents
a missed opportunity for suicide prevention
efforts.13 This is pertinent, as New Zealand has
high levels of alcohol use; 80% of New Zealand-
ers ≥15 years have drunk alcohol in the past year;

New Zealand Medical Journal
Te ara tika o te hauora hapori
2022 Jul 15; 135(1558). ISSN 1175-8716
www.nzma.org.nz/journal ©PMA
article 66
20% drink at hazardous levels.14 However, the
relationship between alcohol and suicide has not
been examined systematically in New Zealand,
and the national Suicide Prevention Strategy fails
to identify alcohol harm reduction strategies as a
means of suicide prevention.15
We sought to inform suicide prevention by
improving understanding of a potentially signif-
icant and modiable risk factor, using New Zea-
land-specic data. The objective of this research
was to quantify and characterise the association
between acute alcohol use and suicide death in
New Zealand, to provide a baseline against which
future interventions or trends can be assessed.
Specically, we asked: 1) how prevalent is acute
alcohol use within suicide deaths in New Zealand;
2) has the proportion of suicides involving acute
alcohol use changed over time, and 3) what are
the characteristics of suicides involving acute
alcohol use?
Methods
Data were drawn from the National Coronial
Information System (NCIS), which compiles cases
from the Coronial Service of New Zealand. This
project was approved by the University of Otago
Human Research Ethics Committee (HD20/102)
with a second level of review and approval by
NCIS (NZ019).
Case identification and inclusion criteria
New Zealand suicide data are available from
NCIS for 1 July 2007 onwards. Cases were extracted
from this date to 31 December 2020, with the
search last run on 7 September 2021. Eligibility
criteria were closed cases, coded in NCIS as inten-
tional self-harm (i.e., suicide), where the person
was ≥15 years at death. The rationale for this age
cut-off recognises that suicidal intent differs in
individuals, and while intent may be determined
in children as young as 12,16 the World Health
Organization uses 15 as the lower age group in
global statistics and reporting. As cases can take
up to two years to be closed, not all suicide deaths
from 2019 and 2020 were included in the sample
of eligible cases. NCIS reports quarterly on case
closure percentages across all deaths,17 based on
information provided by the Coronial Services of
New Zealand; the report for all deaths (not spe-
cic to suicide) that most closely matches the data
collection period of this study was published on
1 October 2021, and it indicated that case closure
percentages for 2019 and 2020 at that time were
70.5% and 55.8%, respectively. Inclusion criteria
were then applied in the following sequence:
The rationale for criterion 4 (above) is that
alcohol can be produced endogenously through
decomposition; however, vitreous humour is less
prone to microbial invasion and post-mortem
effects.18 These criteria are consistent with those
of a recent Australian study,7 and facilitate com-
parison of results.
Characteristics of suicide deaths
The following characteristics were extracted,
primarily using NCIS-coded data with any missing
data searched for in the linked coronial reports:
• Age – age in years at death, subsequently
grouped into ten-year intervals
• Sex – female, male
• Employment status – employed,
unemployed, student, retired/pensioned,
other (including categories of home duties,
prisoner, still enquiring, child not at school),
unknown
• Marital status – Never married, widowed,
divorced/separated, married/de facto,
unknown
• Ethnicity – European, Māori, Pacic peoples,
Asian, Middle Eastern/Latin American/
African, other ethnicity (consistent with
2018 Census ethnic group summaries)19
• Method of death – as follows, based on
International Classication of Diseases ICD-
10-AM code;
• Poisoning – X40-X49, X60-X69, X85-X90,
Y10-Y19
• Hanging – W75-W84, X70, X91, Y20
• Drowning – W65-W74, X71, X92, Y21
• Firearm – W32-W34, X72-X74, X93-X95,
Y22-Y24
• Sharp object – W25-W29, X78, X99, Y28
• Falls – W00-W19, X80, Y01, Y30
1. Is toxicology data available?
2. Was alcohol measured in blood post-mortem
in toxicological analysis?
3. If an additional ante-mortem blood sample
was taken, post-mortem and ante-mortem
samples must concur.
4. If decomposition was noted, alcohol
concentration must also have been
conrmed through measurement in vitreous
humour.

New Zealand Medical Journal
Te ara tika o te hauora hapori
2022 Jul 15; 135(1558). ISSN 1175-8716
www.nzma.org.nz/journal ©PMA
article 67
• Other – W22-W23, X00, X30-X31, X75-
X79, X81-X84, Y23, Y31
• Year of death.
Risk factors such as mental health history,
chronic pain, or nancial problems are not coded
in NCIS. These factors were therefore excluded
from analyses, because it cannot be assumed that
the absence of these factors in Coronial or police
reports means that they were not present.
Post-mortem BAC was extracted from toxicologi-
cal and coronial reports and then dichotomised as:
• No acute alcohol use – BAC ≤50mg/100mL of
blood
• Acute alcohol use – BAC >50mg/100mL of
blood.
This categorisation is consistent with the cur-
rent legal BAC for adults (20 years and older)
when driving in New Zealand (i.e., 0.05%). As
a sensitivity test, analyses were run using two
additional BAC levels to dene acute alcohol use:
>30mg/100mL (where some individuals may show
signs of impairment) and >80mg/100mL (New Zea-
land’s legal driving limit until December 2014).
Whether alcohol was identied in NCIS records
as a contributory cause of death was also deter-
mined by use of ICD-10-AM codes (F10.0, F10.1–
10.9, R78.0, T51, X45 or X65) and searching for the
word “alcohol” at all levels of the cause of death
elds. If any one of these codes was identied, we
deemed alcohol had been identied as a contrib-
utory cause of death. In many instances, blood
alcohol was the only toxicological test on record.
Therefore, data on other psychoactive substances
were not extracted.
Statistical analysis
Statistical analysis was conducted in Stata (ver-
sion 16.1 for Windows).20 Fields with n<5 were
blinded to minimise risk of individuals being
identied, with associated cell counts suppressed
to prevent blinded cells from being calculated.
For each year of the study period, the propor-
tion of suicide deaths involving acute alcohol use
was calculated, to avoid the effects of population
change over that time. Data are presented visu-
ally, without statistical analysis. These data are
included to enable assessment of any future inter-
vention relative to baseline trends.
To test for differences between included and
excluded cases, and to compare characteristics of
suicide deaths with and without acute alcohol use,
logistic regression modelling was undertaken.
Multivariate logistic regression models were used
to compute unadjusted and adjusted odds ratios
(OR, AOR; 95% CIs) controlling for the effects of
all other signicant variables (age, sex, employ-
ment status, marital status, ethnicity and method
of death).
Results
Between 1 July 2007 and 31 December 2020,
6,072 New Zealanders aged ≥15 years died by sui-
cide and had NCIS records. No toxicology data
were available for 651 cases (10.7%); no mea-
surement of BAC was undertaken post-mortem
(n=235; 3.9%); ante-mortem and post-mortem BAC
levels did not concur (n=4; 0.1%); and decompo-
sition was noted but alcohol concentration was
not conrmed in vitreous humour (n=524; 8.6%).
These 1,414 cases were, therefore, excluded leav-
ing a total sample of 4,658 cases (76.7%) eligible
for inclusion. The percentage of cases excluded
per year of the study period are shown in Supple-
mentary Table 1.
Table 1 compares characteristics of excluded
and included cases. There was no signicant dif-
ference between excluded and included cases in
relation to sex. Older age groups were less likely
to be included than those aged 15–24. Those who
were widowed were less likely to be included than
those who were never married, while those who
were married were more likely to be included.
Of included suicides, 1,238 (26.6%) involved
acute alcohol use, BACs are shown in Table 2.
Figure 1 shows the proportion of suicides involv-
ing acute alcohol use across the 14-year study
period. This fraction ranged from 21.7% to 33.3%
across the 14 years (2007–2020), with no clear
trend over time.
Of the 1,238 suicides involving acute alcohol
use, 416 (33.6%) were coded in NCIS as alcohol
being a contributory cause of death (the most com-
monly identied codes were F10.0 (acute alcohol
intoxication) and F10.1 (harmful use of alcohol)).
In addition, 154 suicides where acute alcohol use
was not identied (BAC ≤50mg/100mL) also had
alcohol coded as a contributory cause of death,
occurring predominantly in the context of alcohol
dependence and/or a low level of BAC considered
contributory to a poly-drug overdose.
The characteristics of suicides without and
with acute alcohol use are described in Table 3.
The proportion of suicides involving acute alco-
hol use declined with increasing age; however,

New Zealand Medical Journal
Te ara tika o te hauora hapori
2022 Jul 15; 135(1558). ISSN 1175-8716
www.nzma.org.nz/journal ©PMA
article 68
Table 1: Descriptive characteristics of included cases compared to excluded cases and results of logistic regression modelling showing characteristics associated with inclusion.
Characteristic Excluded n (%) Included n (%) OR (95% CI) AOR (95% CI)
Age 15–24 240 (18.2) 1,081 (81.8) REF REF
25–34 236 (22.0) 836 (78.0) 0.79 (0.64–0.96) * 0.73 (0.59–0.91) **
35–44 239 (23.0) 802 (77.0) 0.75 (0.61–0.91) ** 0.67 (0.54–0.85) **
45–54 279 (23.7) 897 (76.3) 0.74 (0.59–0.87) ** 0.64 (0.51–0.81) ***
55–64 208 (26.7) 571 (73.3) 0.61 (0.49–0.75) *** 0.56 (0.43–0.72) ***
65–74 7 (25.9) 20 (74.1) 0.63 (0.27–1.52) 0.62 (0.25–1.54)
75+ 205 (31.3) 451 (68.8) 0.49 (0.39–0.61) *** 0.61 (0.42–0.90) *
Sex Female 360 (22.9) 1,211 (77.1) REF REF
Male 1,054 (23.4) 3,447 (76.6) 0.97 (0.85–1.11) 0.97 (0.84–1.12)
Employment status Employed 602 (22.3) 2,101 (77.7) REF REF
Unemployed 377 (23.2) 1,248 (76.8) 0.95 (0.82–1.10) 1.01 (0.87–1.17)
Student 83 (18.2) 372 (81.8) 1.28 (1.00–1.66) 1.11 (0.84–1.49)
Rered/
pensioner 240 (32.3) 504 (67.7) 0.60 (0.50–0.72) *** 0.71 (0.52–0.96) *
Other 47 (16.8) 233 (83.2) 1.42 (1.03–1.97) * 1.48 (1.06–2.06) *
Unknown 65 (24.5) 200 (75.5) 0.88 (0.66–1.18) 0.95 (0.71–1.29)
Marital status Never married 610 (24.0) 1,930 (76.0) REF REF
Widowed 82 (34.3) 157 (65.7) 0.61 (0.46–0.80) *** 1.07 (0.77–1.49)
Divorced/
separated 224 (25.8) 646 (74.3) 0.91 (0.76–1.09) 1.20 (0.99–1.46)
Married/
de facto 410 (19.5) 1,695 (80.5) 1.31 (1.13–1.50) *** 1.66 (1.41–1.95) ***
Unknown 88 (27.7) 230 (72.3) 0.83 (0.64–1.07) 0.97(0.74–1.27)

New Zealand Medical Journal
Te ara tika o te hauora hapori
2022 Jul 15; 135(1558). ISSN 1175-8716
www.nzma.org.nz/journal ©PMA
article 69
Characteristic Excluded n (%) Included n (%) OR (95% CI) AOR (95% CI)
Ethnicity European 1,023 (24.4) 3,174 (75.6) REF REF
Māori 258 (20.8) 983 (79.2) 1.23 (1.05–1.43) ** 0.93 (0.78–1.10)
Pacic peoples 61 (23.3) 201 (76.7) 1.06 (0.79–1.43) 0.79 (0.58–1.07)
Asian 62 (19.8) 252 (80.3) 1.31 (0.98–1.74) 1.14 (0.85–1.53)
Middle Eastern/Lan
American/African 5 (19.2) 21 (80.8) 1.35 (0.51–3.60) 1.18 (0.44–3.18)
Other ethnicity 5 (15.6) 27 (84.4) 1.74 (0.67–4.53) 1.74 (0.66–4.62)
Method of death Poisoning 369 (28.4) 932 (71.6) REF REF
Hanging 746 (20.4) 2,911 (79.6) 1.55 (1.34–1.79) *** 1.34 (1.15–1.57) ***
Drowning 33 (28.7) 82 (71.3) 0.98 (0.65–1.50) 1.02 (0.67–1.57)
Firearm 119 (25.0) 358 (75.1) 1.19 (0.94–1.51) 1.13 (0.88–1.45)
Sharp object 38 (31.2) 84 (68.9) 0.86 (0.59–1.31) 0.83 (0.55–1.25)
Falls 38 (22.8) 129 (77.3) 1.34 (0.92–1.97) 1.24 (0.84–1.83)
Other 71 (30.5) 162 (69.5) 0.90 (0.67–1.22) 0.83 (0.61–1.13)
NB: Multivariate logistic regression models were used to compute both unadjusted odds ratios (OR), and adjusted odds ratios (AOR) controlling for the eects of all other significant variables (age, sex, employment
status, marital status, ethnicity and method of death). Data are reported with 95% confidence intervals (CI). REF – reference group. * p<0.05, ** p<0.01, *** p<0.001.
Table 1 (continued): Descriptive characteristics of included cases compared to excluded cases and results of logistic regression modelling showing characteristics associated with inclusion.

New Zealand Medical Journal
Te ara tika o te hauora hapori
2022 Jul 15; 135(1558). ISSN 1175-8716
www.nzma.org.nz/journal ©PMA
article 70
the 65–74 age group data could not be reported
due to small numbers. The fraction of male and
female suicides involving acute alcohol use were
almost identical. Of these, males were 26.5%;
median BAC 142mg/100mL; interquartile range
101–188mg/100mL); and females were 26.7%;
median BAC 155mg/100mL; interquartile range
109–203mg/100mL). Māori (32.3%) and Pacic
peoples (35.3%) had higher proportions of sui-
cides involving acute alcohol use than Europeans
(25.4%) and Asians (11.9%).
Table 4 summarises the logistic regression
model comparing suicide deaths with and with-
out acute alcohol use, and shows ORs for each
independent variable, unadjusted and adjusted
(for age, sex, employment status, marital sta-
tus, ethnicity, suicide means). This analysis con-
rms the association between suicide involving
acute alcohol use, and between being young/
middle aged (<55 years), employed, of Māori or
Pacic ethnicity, and using hanging as the sui-
cide method. Overall, there was minimal effect
on the ORs after adjustment, except for being
widowed (120 (76.4%) of those who were wid-
owed were aged 75+). Sensitivity tests of the two
additional BACs utilised for logistic regression
modelling are presented in the Supplementary
File; there was a minimal impact on the ORs;
however, some covariates showed an alteration
in signicance level.
Table 2: Identied blood alcohol concentrations (BAC) in included suicide deaths.
BAC range n %
≤50mg/100mL 3,420 73.4%
51–100mg/100mL 289 6.2%
101–150 mg/100mL 363 7.8%
151–200 mg/100mL 322 6.9%
mg/100mL 175 3.8%
> 250mg/100mL 89 1.9%
NB: BAC – blood alcohol concentration reported in milligrams per 100 millilitres of blood.
Figure 1: The proportion of suicide deaths with acute alcohol use from 2007–2020 (shown as a percentage by year)
show no consistent increasing or decreasing trend over the study period.

New Zealand Medical Journal
Te ara tika o te hauora hapori
2022 Jul 15; 135(1558). ISSN 1175-8716
www.nzma.org.nz/journal ©PMA
article 71
Table 3: Characteristics comparison between suicide deaths without and with acute alcohol use (AAU).
Characteristic
No AAU
n (%)
AAU
n (%)
Total
Age 15–24 771 (71.3) 310 (28.7) 1,081
25–34 572 (68.4) 264 (31.6) 836
35–44 553 (69.0) 249 (31.0) 802
45–54 656 (73.1) 241 (26.9) 897
55–64 448 (78.5) 123 (21.5) 571
65–74 Blinded for confidentiality n<5 20
75+ 404 (89.6) 47 (10.4) 451
Sex Female 888 (73.3) 323 (26.7) 1,211
Male 2,532 (73.5) 915 (26.5) 3,447
Employment status Employed 1,439 (68.5) 662 (31.5) 2,101
Unemployed 911 (73.0) 337 (27.0) 1,248
Student 296 (79.6) 76 (20.4) 372
Retired/pensioner 448 (88.9) 56 (11.1) 504
Other 185 (79.4) 48 (20.6) 233
Unknown 141 (70.5) 59 (29.5) 200
Marital status Never married 1,393 (72.2) 537 (27.8) 1,930
Widowed 125 (79.6) 32 (20.4) 157
Divorced/separated 467 (72.3) 179 (27.7) 646
Married/de facto 1,272 (75.0) 423 (25.0) 1,695
Unknown 163 (70.9) 67 (29.1) 230

New Zealand Medical Journal
Te ara tika o te hauora hapori
2022 Jul 15; 135(1558). ISSN 1175-8716
www.nzma.org.nz/journal ©PMA
article 72
Characteristic
No AAU
n (%)
AAU
n (%)
Total
Ethnicity European 2,368 (74.6) 806 (25.4) 3,174
Māori 665 (67.7) 318 (32.3) 983
Pacific peoples 130 (64.7) 71 (35.3) 201
Asian 222 (88.1) 30 (11.9) 252
Middle Eastern/
Latin American/African Blinded for confidentiality n<5 21
Other ethnicity 17 (63.0) 10 (37.0) 27
Method of death Poisoning 722 (77.5) 210 (22.5) 932
Hanging 2,044 (70.2) 867 (29.8) 2,911
Drowning 67 (81.7) 15 (18.3) 82
Firearm 272 (76.0) 86 (24.0) 358
Sharp object 71 (84.5) 13 (15.5) 84
Falls 110 (85.3) 19 (14.7) 129
Other 134 (82.7) 28 (17.3) 162
NB: Results are not shown when numbers are less than five (n<5) and related cells are blinded to prevent calculation of suppressed cells (which would increase the likelihood that individual cases
could be identified).
Table 3 (continued): Characteristics comparison between suicide deaths without and with acute alcohol use (AAU).

New Zealand Medical Journal
Te ara tika o te hauora hapori
2022 Jul 15; 135(1558). ISSN 1175-8716
www.nzma.org.nz/journal ©PMA
article 73
Table 4: Summary of logistic regression modelling for risk of suicide death involving acute alcohol use (AAU) by case characteristics.
Characteristic OR 95% CI p value AOR 95% CI p value
Age 15–24 REF REF
25–34 1.15 0.94, 1.40 0.169 1.11 0.90, 1.38 0.325
35–44 1.12 0.92, 1.37 0.266 1.11 0.88, 1.40 0.370
45–54 0.91 0.75, 1.11 0.371 0.93 0.73, 1.19 0.575
55–64 0.68 0.54, 0.87 0.002 0.74 0.56, 0.99 0.041
65–74 0.62 0.21, 1.87 0.399 0.90 0.29, 2.86 0.863
75+ 0.29 0.21, 0.40 <0.001 0.42 0.25, 0.72 0.002
Sex Female REF REF
Male 0.99 0.86, 1.15 0.931 0.87 0.74, 1.02 0.087
Employment Employed REF REF
Unemployed 0.80 0.69, 0.94 0.006 0.73 0.62, 0.85 <0.001
Student 0.56 0.43, 0.73 <0.001 0.49 0.36, 0.66 <0.001
Retired/pensioner 0.27 0.20, 0.36 <0.001 0.49 0.31, 0.76 0.002
Other 0.56 0.41, 0.78 0.001 0.51 0.36, 0.71 <0.001
Unknown 0.91 0.66, 1.25 0.559 0.86 0.62, 1.20 0.375
Marital status Never married REF REF
Widowed 0.66 0.44, 0.99 0.045 1.79 1.10, 2.90 0.019
Divorced/separated 0.99 0.81, 1.21 0.955 1.05 0.84, 1.31 0.654
Married/de facto 0.86 0.74, 1.00 0.051 0.92 0.77, 1.10 0.329
Unknown 1.07 0.79, 1.44 0.676 1.19 0.86, 1.63 0.289

New Zealand Medical Journal
Te ara tika o te hauora hapori
2022 Jul 15; 135(1558). ISSN 1175-8716
www.nzma.org.nz/journal ©PMA
article 74
Characteristic OR 95% CI p value AOR 95% CI p value
Ethnicity European REF
Māori 1.40 1.20, 1.64 <0.001 1.20 1.01, 1.42 0.043
Pacific peoples 1.60 1.19, 2.17 0.002 1.46 1.10, 2.00 0.018
Asian 0.40 0.27, 0.59 <0.001 0.42 0.28, 0.63 <0.001
Middle Eastern/Latin
American/African 0.49 0.14, 1.67 0.253 0.45 0.13, 1.56 0.208
Other ethnicity 1.73 0.79, 3.79 0.172 1.68 0.75, 3.77 0.209
Method of death Poisoning REF REF
Hanging 1.46 1.23, 1.73 <0.001 1.30 1.07, 1.56 0.007
Drowning 0.77 0.43, 1.38 0.377 1.04 0.57, 1.89 0.891
Firearm 1.09 0.82, 1.45 0.569 1.10 0.81, 1.48 0.544
Sharp object 0.63 0.34, 1.16 0.138 0.65 0.35, 1.21 0.174
Falls 0.59 0.36, 0.99 0.045 0.58 0.34, 0.98 0.040
Other 0.72 0.46, 1.11 0.136 0.67 0.43, 1.05 0.078
NB: Multivariate logistic regression models were used to compute both unadjusted odds ratios (OR), and adjusted odds ratios (AOR) controlling for the eects of all other significant variables (age, sex, employment
status, marital status, ethnicity and method of death). Data are reported with 95% confidence intervals (CI). REF – reference group.
Table 4 (continued): Summary of logistic regression modelling for risk of suicide death involving acute alcohol use (AAU) by case characteristics.

New Zealand Medical Journal
Te ara tika o te hauora hapori
2022 Jul 15; 135(1558). ISSN 1175-8716
www.nzma.org.nz/journal ©PMA
article 75
Discussion
We quantied and characterised the associa-
tion between acute alcohol use and suicide in New
Zealand, by analysing coronial data from 2007 to
2020, in order to provide a baseline dataset for
the association. We found that around one quar-
ter (26.6%) of all suicides over the study period
involved acute alcohol use. While this is the rst
time that the proportion of suicide deaths involv-
ing acute alcohol use has been quantied in New
Zealand, these ndings are consistent with inter-
national studies: meta-analytic ndings show the
prevalence of acute alcohol use in suicides inter-
nationally ranged from 26.5% to 44.4%,2 and a
recent Australian study found 26.7% of suicides
between 2010 and 2015 involved acute alcohol
use.7 The proportion of New Zealand suicides
involving acute alcohol use was stable over the
14-year study period, indicating that acute alco-
hol use has a strong, persistent and long-standing
association with suicide.
While the overall proportion of suicides involv-
ing acute alcohol use in New Zealand was com-
parable with other similar countries, we found
key differences in demographic characteristics.
Equal fractions of male and female suicides in
New Zealand involved acute alcohol use, in con-
trast to international studies which have consis-
tently identied males as having a higher fraction
of suicides involving acute alcohol use.7–9 This
difference may reect New Zealand’s alcohol
culture, but is dicult to explain given that data
from the New Zealand Health Survey shows that
hazardous drinking rates in males are approxi-
mately double those of females.14 However, New
Zealand has rates of foetal alcohol spectrum dis-
order substantially higher than the global prev-
alence estimate,21 which suggests that female
alcohol consumption may be higher than found
in current data sources. In a number of OECD
countries, including Australia and New Zealand,
suicides in young females have increased, partic-
ularly among young indigenous females.22 Given
that risk factors for binge drinking differ between
males and females,23 there is a need for further
research focussed on female alcohol use, ethnic-
ity, drinking patterns and suicidal behaviour, in
order to inform development of interventions
specic to female needs, and which are culturally
appropriate and responsive.
We found that those aged between 15 and 54
years had similar risks of suicide involving acute
alcohol use, in contrast to the Australian study
that found middle age groups (ages 35–44) had
increased risk.7 This nding is of concern given
New Zealand’s high teenage suicide rate,24 and
points to alcohol use being an important point
of intervention in reducing teenage suicide.
However, these ndings were pooled across the
study period, and adolescent hazardous drinking
declined overall in New Zealand between 2001
and 2012,25 while hazardous drinking in older peo-
ple is of increasing concern.26 These time-dynamic
changes suggest the need to monitor consumption
patterns in different demographic groups, as well
as the relationship between acute alcohol use and
suicide. We also found signicant ethnicity differ-
ences, with Māori and Pacic peoples more likely
to die by suicide involving acute alcohol use than
European and Asian ethnicities. This observation
is a substantial health equity issue, and may reect
multiple risk factors for hazardous alcohol use
that disproportionately impact Māori and Pacic
peoples, including; neighbourhood availability
of alcohol,27 experiences of discrimination,28 and
the effects of trauma.29 There are well-established
inequities in New Zealand for both alcohol-related
harm30 and suicide;10 this study adds to that body
of knowledge and supports the need for address-
ing alcohol as a contributor to health inequities.
This study provides the rst known quantica-
tion of acute alcohol use in New Zealand suicides,
but limitations need to be acknowledged. Some
differences in the characteristics of included and
excluded cases may be a potential source of bias.
In particular, older age groups were more likely
to be excluded from the nal sample of cases,
with almost half of exclusions due to a lack of
toxicology data. Reasons for this are unclear but
the characteristics of the decedent may inuence
testing decisions and contribute to possible bias.
Additionally, BAC was the only toxicological test
ordered in many cases, which means that the
contribution of other psychoactive substances
could not be evaluated. We recommend that tox-
icology should be ordered, and BAC analysed, for
every suspected suicide. This recommendation is
important, since NCIS-coding of alcohol as a con-
tributory cause of death cannot be solely relied
upon as a denition of alcohol’s contribution to
suicide. We found that only one third of suicides
involving acute alcohol use (as dened by BAC
results) in New Zealand had alcohol coded as con-
tributory only. A comparable Australian study
found that half of suicides involving acute alcohol
use had alcohol coded as a contributory cause of
death.31 We acknowledge that dening acute alco-

New Zealand Medical Journal
Te ara tika o te hauora hapori
2022 Jul 15; 135(1558). ISSN 1175-8716
www.nzma.org.nz/journal ©PMA
article 76
hol use at a set BAC does not reect that alcohol’s
effects may differ by individual. However, our
approach was consistent with previous studies,7,8
and our sensitivity testing suggests that the pres-
ence of alcohol is more relevant than the cut-off
used to dene acute alcohol use. Results are cur-
rent, as at 7 September 2021, acknowledging that
cases may still be closed and added after this date
given that there is a lag between year of death
and coronial cases being closed, particularly for
the years of 2019 and 2020. As such, it would be
benecial to re-run this study a few years in the
future, to add to long-term trends, and to identify
any pandemic-related impacts.
The design of this study does not allow consid-
eration of the mechanism of association between
acute alcohol use and suicide,3 nor can we deter-
mine whether alcohol was used as a deliberate,
facilitatory means of suicide.32 Another limitation is
that not all relevant variables are consistently avail-
able within the NCIS dataset e.g., socio-economic
status, co-morbid mental disorders (particularly
alcohol use disorder and other substance use disor-
ders), and acute stressors prior to death. Based on
international ndings, we hypothesise that those
who die by suicide involving acute alcohol use are
more likely to die impulsively following acute stress-
ful events, rather than having psychiatric or physi-
cal co-morbidities.8,9 Future New Zealand research
should investigate prior service contact for those
who die by suicide involving acute alcohol use.
We have identied that just over a quarter
of suicide deaths in New Zealand involve acute
alcohol use, and this is particularly prevalent in
population groups known to have higher suicide
rates (including young people and Māori). Thus,
we conclude that alcohol use is a signicant but
modiable risk factor for a substantial group of
suicide deaths in New Zealand. International
evidence shows that actions taken to reduce
alcohol consumption at a population-level are
associated with reduced suicidal behaviour.33
Therefore, we recommend that interventions
targeted at alcohol be included in New Zealand’s
suicide prevention strategy. Our ndings pro-
vide baseline data for the development of inter-
ventions targeting suicide associated with acute
alcohol use in New Zealand.
Appendix:
https://uploads-ssl.webow.com/5e332a62c703f6340a2faf44/62ccecc35b92a363774f0311_5693%20-%20appendix-nal.pdf

New Zealand Medical Journal
Te ara tika o te hauora hapori
2022 Jul 15; 135(1558). ISSN 1175-8716
www.nzma.org.nz/journal ©PMA
article 77
 
Nil.

This research was funded by the Oakley Mental Health
Foundation (grant awarded to Rose Crossin and
Joseph M Boden). Author Katrina Witt is funded by an
emerging leader fellowship from the National Health
and Medical Research Council of Australia (1177787).
We acknowledge the Victorian Department of Justice
and Community Safety as the organisation source of
the data, and the National Coronial Information System
(NCIS) as the database source of the data. The authors
thank the sta of the National Coronial Information
System for their assistance throughout this project,
including providing feedback on the dra manuscript.
 
Rose Crossin: Department of Population Health,
University of Otago (Christchurch), Christchurch,
New Zealand.
Lana Cleland: Department of Population Health,
University of Otago (Christchurch), Christchurch,
New Zealand; Department of Psychological Medicine,
University of Otago (Christchurch), Christchurch,
New Zealand.
Annette Beautrais: South Canterbury District Health
Board, Timaru, New Zealand.
Katrina Witt: Orygen, Parkville, Australia; Centre for
Youth Mental Health, The University of Melbourne,
Parkville, Australia.
Joseph M Boden: Department of Psychological Medicine,
University of Otago (Christchurch), Christchurch,
New Zealand.
 
Rose Crossin: Department of Population Health,
University of Otago (Christchurch), Christchurch,
New Zealand. rose.crossin@otago.ac.nz

www.nzma.org.nz/journal-articles/acute-alcohol-
use-and-suicide-deaths-an-analysis-of-new-zealand-
coronial-data-from-2007-2020

1. Turecki G, Brent DA, Gunnell D, O’Connor RC,
Oquendo MA, Pirkis J, et al. Suicide and suicide risk.
Nature reviews Disease primers. 2019;5(1):1-22.
2. Borges G, Bagge C, Cherpitel CJ, Conner K, Orozco
R, Rossow I. A meta-analysis of acute alcohol use
and the risk of suicide attempt. Psychological
medicine. 2017;47(5):949.
3. Cherpitel CJ, Borges GL, Wilcox HC. Acute alcohol
use and suicidal behavior: a review of the literature.
Alcoholism: clinical and experimental research.
2004;28:18S-28S.
4. Sher L. Alcohol consumption and suicide. QJM.
2006;99(1):57-61.
5. Huord MR. Alcohol and suicidal behavior. Clinical
psychology review. 2001;21(5):797-811.
6. Norström T, Rossow I. Alcohol consumption as
a risk factor for suicidal behavior: a systematic
review of associations at the individual and at the
population level. Archives of Suicide Research.
2016;20(4):489-506.
7. Chong D, Buckley N, Schumann J, Chitty K.
Acute alcohol use in Australian coronial suicide
cases, 2010-2015. Drug and alcohol dependence.
2020:108066.
8. Lee JW, Park CHK, Kim EY, Kim SH, Yoo SH, Ahn YM.
Characteristics of completed suicide in dierent
blood alcohol concentrations in Korea. Forensic
science international. 2017;281:37-43.
9. Kõlves K, Koo YW, De Leo D. A drink before suicide:
analysis of the Queensland Suicide Register in
Australia. Epidemiology and psychiatric sciences.
2020;29.
10. Coronial Services of New Zealand. Annual Suicide
Statistics. Wellington: Ministry of Justice; 2021.
11. World Health Organization. Global status report on
alcohol and health 2018. World Health Organization;
2019. Report No.: 9241565632.
12. World Health Organisation. Preventing suicide:
a global imperative. Geneva: World Health
Organization; 2014.
13. Witt K, Lubman DI. Eective suicide prevention:
Where is the discussion on alcohol? Australian
& New Zealand Journal of Psychiatry.
2018;52(6):507-8.
14. Ministry of Health. New Zealand Health
Survey 2019/20 Wellington: Ministry of Health;
2019 [Available from: https://minhealthnz.
shinyapps.io/nz-health-survey-2018-19-
annual-data-explorer/_w_da15d6ce/#!/
explore-indicators.
15. Ministry of Health. Every Life Matters - He Tapu te
Oranga o ia tangata: Suicide Prevention Strategy
2019-2029 and Suicide Prevention Action Plan
2019-2024 for Aotearoa New Zealand. Wellington:
Ministry of Health; 2019.
16. Ridge Anderson A, Keyes GM, Jobes DA.
Understanding and treating suicidal risk in young
children. Practice Innovations. 2016;1(1):3.
17. National Coronial Information System. NCIS
Operational Statistics Australia: National Coronial
Information System; 2022 [updated April 2022; cited
2022 3 May 2022]. Available from: https://www.ncis.

New Zealand Medical Journal
Te ara tika o te hauora hapori
2022 Jul 15; 135(1558). ISSN 1175-8716
www.nzma.org.nz/journal ©PMA
article 78
org.au/about-the-data/operational-statistics/.
18. Kugelberg FC, Jones AW. Interpreting results
of ethanol analysis in postmortem specimens:
a review of the literature. Forensic science
international. 2007;165(1):10-29.
19. Statistics New Zealand. 2018 Census ethnic group
summaries Wellington, New Zealand2018 [cited
2021. Available from: https://www.stats.govt.nz/
tools/2018-census-ethnic-group-summaries.
20. StataCorp. Stata Statistical Soware: Release 16. .
College Station, TX: StataCorp LLC.; 2019.
21. Lange S, Probst C, Gmel G, Rehm J, Burd L, Popova
S. Global prevalence of fetal alcohol spectrum
disorder among children and youth: a systematic
review and meta-analysis. JAMA pediatrics.
2017;171(10):948-56.
22. Stefanac N, Hetrick S, Hulbert C, Spittal MJ, Witt K,
Robinson J. Are young female suicides increasing? A
comparison of sex-specific rates and characteristics
of youth suicides in Australia over 2004–2014. BMC
Public Health. 2019;19(1):1-11.
23. Wilsnack RW, Wilsnack SC, Gmel G, Kantor LW.
Gender dierences in binge drinking: Prevalence,
predictors, and consequences. Alcohol research:
current reviews. 2018.
24. UNICEF Oice of Research. Building the future:
Children and the sustainable development goals in
rich countries. Innocenti, Florence: UNICEF Oice of
Research 2017.
25. Ball J, Edwards R, Sim D, Cook H, Denny S. What
explains the decline in adolescent binge-drinking in
New Zealand? International Journal of Drug Policy.
2020;84:102826.
26. Stevenson BS, Stephens C, Dulin P, Kostick M,
Alpass F. Alcohol consumption among older adults
in Aotearoa/New Zealand: a comparison of ‘baby
boomers’ and ‘over-65s’. Health Psychology and
Behavioral Medicine. 2015;3(1):366-78.
27. Ayuka F, Barnett R, Pearce J. Neighbourhood
availability of alcohol outlets and hazardous alcohol
consumption in New Zealand. Health & Place.
2014;29:186-99.
28. Winter T, Riordan BC, Surace A, Scarf D. Association
between experience of racial discrimination and
hazardous alcohol use among Māori in Aotearoa
New Zealand. Addiction. 2019;114(12):2241-6.
29. Reid J, Taylor-Moore K, Varona G. Towards a
social-structural model for understanding current
disparities in Māori health and well-being. Journal
of loss and trauma. 2014;19(6):514-36.
30. Walker K. Issues of tobacco, alcohol and other
substance abuse for Māori. Wellington; 2019.
31. Chitty KM, Schumann JL, Moran LL, Chong DG,
Hurzeler TP, Buckley NA. Reporting of alcohol as a
contributor to death in Australian national suicide
statistics and its relationship to post-mortem
alcohol concentrations. Addiction. 2020.
32. Conner KR, Bagge CL, Goldston DB, Ilgen MA.
Alcohol and suicidal behavior: what is known and
what can be done. American journal of preventive
medicine. 2014;47(3):S204-S8.
33. Kõlves K, Chitty KM, Wardhani R, Värnik A, De
Leo D, Witt K. Impact of alcohol policies on
suicidal behavior: a systematic literature review.
International journal of environmental research and
public health. 2020;17(19):7030.