Self-poisoning suicide deaths in people with bipolar disorder: characterizing a subgroup and identifying treatment patterns
Schaffer et al. Int J Bipolar Disord
Self-poisoning suicide deaths in people with bipolar disorder: characterizing a subgroup and identifying treatment patterns
Ayal Schaffer 0 2 3
Lauren M. Weinstock 1
Mark Sinyor 2 7
Catherine Reis 0 3
Benjamin I. Goldstein 5 6
Lakshmi N. Yatham 4
Anthony J. Levitt 2 7
0 Mood and Anxiety Disorders Program, Department of Psychiatry, Sunnybrook Health Sciences Centre , 2075 Bayview Avenue, Room FG 52, Toronto, ON M4N 3M5 , Canada
1 Department of Psychiatry and Human Behavior, Brown University , Providence, RI , USA
2 Department of Psychiatry, University of Toronto , Toronto , Canada
3 Mood and Anxiety Disorders Program, Department of Psychiatry, Sunny- brook Health Sciences Centre , 2075 Bayview Avenue, Room FG 52, Toronto, ON M4N 3M5 , Canada
4 Department of Psychiatry, University of British Columbia , Vancouver , Canada
5 Departments of Psychiatry and Pharmacology, University of Toronto , Toronto , Canada
6 Centre for Youth Bipolar Disorder, Depart- ment of Psychiatry, Sunnybrook Health Sciences Centre , Toronto , Canada
7 Department of Psychiatry, Sunnybrook Health Sciences Centre , Toronto , Canada
Objective: To characterize self-poisoning suicide deaths in BD compared to other suicide decedents. Methods: Extracted coroner data from all suicide deaths (n = 3319) in Toronto, Canada from 1998 to 2012. Analyses of demographics, clinical history, recent stressors, and suicide details were conducted in 5 subgroups of suicide decedents: BD self-poisoning, BD other methods, non-BD self-poisoning, non-BD other methods, and unipolar depression self-poisoning. Toxicology results for lethal and present substances were also compared between BD and non-BD self-poisoning subgroups as well as between BD and unipolar depression self-poisoning subgroups. Results: Among BD suicide decedents, self-poisoning was significantly associated with female sex, past suicide attempts, and comorbid substance abuse. In both the BD and non-BD self-poisoning groups, opioids were the most common class of lethal medication. For both groups, benzodiazepines and antidepressants were the most common medications present at time of death, and in 23% of the BD group, an antidepressant was present without a mood stabilizer or antipsychotic. Only 31% of the BD group had any mood stabilizer present, with carbamazepine being most common. No antidepressant, mood stabilizer, or antipsychotic was present in 15.5% of the BD group. Relative to unipolar depression self-poisoning group, the BD self-poisoning group evidenced higher proportion of previous suicide attempt(s) and psychiatry/ER visits in the previous week. Conclusion: People with BD who die by suicide via self-poisoning comprise a distinct but understudied group. The predominant absence of guideline-concordant pharmacologic care comprises a crucial target for future policy and knowledge translation efforts.
Bipolar disorder; Suicide death; Self-poisoning; Overdose; Lethal medications
Suicide is one of the most common causes of death for
people with bipolar disorder (BD) (Angst et al. 2005;
Høyer et al. 2000; Ösby et al. 2001; Pompili et al. 2013a).
Persons with BD account for up to 10% of all suicide
deaths (Chen et al. 2009; Clements et al. 2013; Ilgen et al.
2010; Karch et al. 2006; Schaffer et al. 2014; Takizawa
2012), and have an estimated standardized mortality ratio
of 10–30 compared to the general population (Crump
et al. 2013; Harris and Barraclough 1997; Kessing et al.
2005; Pompili et al. 2013a).
Suicide by self-poisoning is a prevalent cause of death
globally. Self-poisoning accounts for approximately a
quarter of all suicides in England with females and the
young being particularly vulnerable to this method of
suicide (Camidge et al. 2003; Kapur et al. 2005).
Methods of suicide deaths in BD vary across studies and
countries, but self-poisoning deaths are consistently found to
be the first or second most common method (Chen et al.
2009; Dennehy et al. 2011; Gos et al. 2009; Hunt et al.
2006; Isomets and Henriksson 1994; Keks et al. 2009;
© The Author(s) 2017. This article is distributed under the terms of the Creative Commons Attribution 4.0 International License
(http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium,
provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license,
and indicate if changes were made.
Ösby et al. 2001; Rihmer et al. 1990; Schaffer et al. 2014).
Self-poisoning accounted for 25–30% of deaths in the
three largest studies that reported on BD suicide
methods (Chen et al. 2009; Hunt et al. 2006; Ösby et al. 2001)
and 17–53% in the smaller studies (Angst et al. 2005;
Clements et al. 2013; Crump et al. 2013; Harris and
Barraclough 1997; Høyer et al. 2000; Ilgen et al. 2010; Karch
et al. 2006; Kessing et al. 2005; Pompili et al. 2013a;
Schaffer et al. 2014; Takizawa 2012); however, there is
a paucity of data specifically focusing on
characterizing this subgroup of suicide deaths in BD. Furthermore,
BD also presents as a vulnerable psychiatric population
to self-poisoning with elevated prevalence among BD
decedents relative to non-BD decedents (33.5 vs. 17.4%;
Schaffer et al. 2014) and schizophrenia decedents (34.9
vs. 17.4%; Sinyor et al. 2015). Little is known about who
is more likely to die from this method, which substances
are ingested around the time of death, and which of these
specific substances were ultimately lethal. In a small
Finnish study examining 11 people with BD who died
by self-poisoning suicide, lethal ingestion of an
antipsychotic occurred in 45% of cases, followed by 18% with
each of a tricyclic antidepressant, lithium, or
combination of benzodiazepines (Isomets and Henriksson 1994).
There is a modest amount of data on treatments received
near the time of death. In an Australian review of 35 BD
suicide deaths by any method, a clinical panel assessed
only 34% of decedents as having received adequate
therapeutic interventions (pharmacological and psychological)
prior to death (Keks et al. 2009). For instance, while 43%
of patient had taken lithium within 4 weeks of deaths, only
11% had adequate therapeutic lithium levels.
A similar pattern of underutilization of adequately
dosed pharmacotherapy was found in a Finnish sample of
31 BD suicides (Isomets and Henriksson 1994). We are
not aware of any prior studies that extensively report on
the presence of specific classes of medications based on
toxicology at the time of death, a more robust approach
that ensures that the compound was in fact being
ingested. With self-poisoning deaths being common and
a method of suicide that physicians as prescribers have
some control over, it is important to better characterize
this BD subgroup in order to understand who is more
likely to die from this method, and what medications or
other substances were taken around the time of death.
The objective of this study was to examine these data in a
large sample of suicide decedents with BD and
The Office of the Chief Coroner for Ontario (OCC)
investigates all suicide deaths in Toronto, Canada. As part of
a large study of OCC data on all suicides in the City of
Toronto from 1998 to 2012 (n = 3319 suicide deaths)
(Schaffer et al. 2014; Sinyor et al. 2012, 2014), we
examined 5 subgroups: (1) BD suicide by self-poisoning; (2)
BD suicide by other methods; (3) non-BD suicide by
selfpoisoning; (4) non-BD, non-self-poisoning; (5) unipolar
depression by self-poisoning.
OCC charts include a coroner’s investigation report,
pathology report, toxicology report, and collateral
information gathered from interviews with family or others,
physician/clinical records, police reports, and copies of
suicide notes. OCC data are not available for
approximately 2 years after the death while investigations are
completed. Deaths were classified as suicides based on
a standard of a high degree of probability. We did not
include deaths that were considered indeterminate with
respect to suicide as the cause.
A standardized data extraction procedure was used,
collecting data on: (1) demographics: age, sex, marital
status, living circumstances; (2) clinical variables: the
presence of a BD diagnosis, unipolar depression,
substance abuse history (including alcohol, drugs or both),
past suicide attempts, reported contact with psychiatric
or emergency services in the week prior to death,
comorbid medical condition; (3) recent stressors: employment/
financial, interpersonal stressor, medical/health, police/
legal, bereavement; and (4) details of suicide: method,
presence of a suicide note and location of suicide.
Demographic data and details of the suicide were
available in 99% of coroner charts. Clinical and stressor
variables were only included in the coroner charts if they were
present. All clinical and stressor variables were therefore
considered estimates, with possible underreporting.
For the purpose of this study, any person with BD who
died by intentional self-poisoning, as a single or
combination method of suicide, was included in the BD
selfpoisoning suicide group. All other BD suicide deaths
were included in a non-self-poisoning suicide group. As
additional comparison groups, we also included
self-poisoning and non-self-poisoning suicide deaths in people
without BD as well as a unipolar depression
A detailed toxicology report was available for most
deaths by self-poisoning (93.4% of BD group and 85.1%
of the non-BD group). The OCC primarily conducts
toxicology tests in cases where self-poisoning is
suspected, therefore no toxicology data are available for
the group of BD suicides by other methods. Toxicology
reports include information on the level and
distribution of all substances detected in the blood or urine, and
the approximate time since death. The pathologist uses
this information to determine whether the substance
was a cause of death or only present at the time of the
death. Such determinations are complex, as they can
depend on the deceased’s medical history as well as
substance levels that are influenced by the timing of bodily
fluid retrieval. We relied exclusively on the pathologists’
determination of the presence and/or lethality of
specific substances in each case rather than arriving at our
own conclusions from review of raw toxicology data.
We recorded the presence of all substances and whether
they were determined by the pathologist to be a cause of
death (i.e., at lethal levels). These included psychotropic
medications, non-psychotropic medications, over the
counter medications, alcohol, illicit substances, and
Using the same methodology which was previously
published by our group (Schaffer et al. 2014), the presence of
a diagnosis of BD was established based on information
in the coroner’s investigation report. Diagnostic
information was obtained by the coroner from a variety of
sources, including medical records from the decedent’s
physician(s), collateral information from family, police
report of personal documents, and content of the
suicide note that stated a diagnosis of BD or the presence of
depressive symptoms in the absence of BD. This ‘unipolar
depression’ group is comprised likely of individuals who
fit into a variety of depressive conditions (Sinyor et al.
2015). Specific symptom criteria or psychological
autopsies were not available. All suicide deaths with BD or
unipolar depression were included in the analysis.
Comparisons of subgroups were conducted by
univariate analyses (t-tests or χ2) of all demographic, clinical,
stressor, and suicide-specific variables. In order to identify
independent contribution to variance in the presence or
absence of a self-poisoning method of suicide, multivariate
logistic regression was completed, including all pre-death
variables with a p < .1 identified in the univariate analyses.
Differences in the proportion of different medication
classes or specific agents being present or lethal at the
time of death were compared between the BD and
nonBD as well as BD and unipolar depression self-poisoning
groups using a series of χ2 tests.
Ethical approval and privacy
The OCC granted approval to this study and provided
full access to their records for the purposes of completing
this study. The study was approved by the Research Ethics
Board at Sunnybrook Health Sciences Centre, Toronto,
Canada. Strict privacy procedures utilized by the OCC
were fully adhered to, with all extracted data maintained
in an encrypted and de-identified format.
During the study period, 207 people with BD died by
suicide, out of a total of 3319 suicide deaths. Among those
with BD, 76 (36.7%) died by self-poisoning. There were also
585 self-poisoning suicide deaths in people that did not
have BD. Table 1 characterizes 4 groups (BD self-poisoning,
BD non-self-poisoning, non-BD self-poisoning, non-BD
non-self-poisoning) with regards to demographics, other
clinical history, recent stressors, and suicide details. There
were 1764 people with unipolar depression who died by
suicide, of which, 409 (23.2%) died by self-poisoning.
BD self‑poisoning vs. BD other methods
As shown in Table 1, those with BD who died by
self-poisoning as compared to other methods were significantly
more likely to be female, to have made a prior suicide
attempt, and to have died at home. There were
nonsignificant trends towards the BD self-poisoning group
being older, and more likely to have comorbid substance
abuse or comorbid medical condition.
Multivariate logistic regression (Table 2) found
several variables to be independently associated with
selfpoisoning being the method of suicide among people
with BD, including female sex, past suicide attempts, and
comorbid substance abuse.
BD self‑poisoning vs. unipolar depression self‑poisoning
The BD self-poisoning group compared to the
unipolar depression self-poisoning group had a significantly
higher proportion of past suicide attempt(s) (63.2 vs.
49.4%, χ2 (1) = 4.87, p < .05), proportion of individuals
with a psychiatry/ER visit(s) in the previous week (14.5
vs. 6.8%, χ2 (1) = 5.04, p < .05), and proportion of adult
age decedents (92.1 vs. 78.5%, χ2 (1), 7.61, p < .05). In
contrast, there was a significantly higher proportion of older
adults (18.1 vs. 3.9%, χ2 (1) = 9.60, p < .01), individuals
with a recent medical health stressor (14.9 vs. 2.6%, χ2
(1) = 8.56, p < .01), and individuals with any stressor
present (51.8 vs. 36.8%, χ2 (1) = 5.76, p < .05) among those
with unipolar depression who died by self-poisoning
relative to those with BD who died by self-poisoning. Welch–
Satterthwaite t test indicated that those with unipolar
depression who died by self-poisoning (mean age 50.9
years; SD = 15.4) were significantly older than those with
BD who died by self-poisoning (mean age = 46.1 years;
SD = 11.3) (t (132.59) = −3.09, p < .01).
Lethal substances/medications in BD and non‑BD
The specific substances and classes of medications that
were present at lethal levels at the time of self-poisoning
suicide are shown in Fig. 1. In both the BD and non-BD
d re s
Table 2 Regression model of variables associated
with self-poisoning among suicide decedents with bipolar
disorder (n = 207)
Odds ratio 95% CI
All pre-death variables with p < .1 in univariate analyses were included
An additional regression examining association and interaction of diagnosis (BD
or non-BD) and sex with self-poisoning method was conducted
A non-significant interaction term (p = .083) was found
groups, opioids were the most common class of lethal
medication. There was a non-significant trend of higher
likelihood of lethal levels of opioids among BD decedents
(excluding n = 21 with unknown multiple drug toxicity)
who had comorbid substance abuse compared to no such
comorbidity (47.1 vs. 22.9%; χ2 = 3.15, df = 1, p = .076).
There was no difference in identification of lethal opioids
based on substance abuse comorbidity in the non-BD
group (45.2 vs. 39.8%; χ2 = 1.26, df = 1, p = .26).
For the BD group, opioids were followed by
benzodiazepines, antidepressants, and antipsychotics as the next
most common classes of identified lethal medication
taken. Nearly a third of the BD group had multiple drug/
alcohol toxicity as the cause of death.
There were no significant differences in frequency of
opioids, OTC medications, or alcohol between BD and
non-BD groups. The mean number of lethal substances
taken by people with BD who died by self-poisoning
(1.76, SD = 1.2) was significantly higher than the number
taken by people without BD (1.47, SD = 1.03) (t = 1.86,
df = 446, p = .064).
An psycho c
Tricyclic An depressant
non - TCA An depressants
Over the Counter Medica on
Mul ple Drug/Alcohol Toxicity
Bipolar Disorder (n = 71)
Non-Bipolar Disorder (n = 498)
Lethal substances/medications in BD and unipolar
depression self‑poisoning groups
A significantly higher proportion of lethal
antipsychotics (21.2 vs. 8.8%, χ2 (1) = 6.95, p < .01) and mood
stabilizers (17.3 vs. .7%, Fisher’s Exact Test, p < .001) was
noted among those with BD who died by self-poisoning
relative to those with unipolar depression who died by
self-poisoning. There was a significantly higher
proportion of lethal TCA ingestion among those with unipolar
depression who died by self-poisoning relative to those
with BD who died by self-poisoning (23.4 vs. 9.6%, %, χ2
(1) = 4.99, p < .05). The mean total number of lethal
substances was not significantly different between those with
BD who died by self-poisoning (mean = 1.76; SD = 1.23)
and those with unipolar depression who died by
self-poisoning (mean = 1.52; SD = 1.04) groups, p > .05.
Substances/medications present in BD and non‑BD
Figure 2 displays the substances present at the time of
death among BD and non-BD self-poisoning suicide
decedents. For the BD group, benzodiazepines were the
most common type of medication present at time of
death (62%), followed closely by antidepressants (58%).
Only 31% of those with BD had any type of traditional
mood stabilizer detected, of which carbamazepine (14%)
was the most common. Thirty-nine percent had
detectable alcohol and had 14% had an illegal substance present.
Thirty-seven percent of the BD group had some level of
opioid present at the time of death. There was a
nonsignificant trend of higher likelihood of opioids being
present among BD decedents with comorbid substance
abuse compared to no such comorbidity (52.2 vs. 29.2%;
χ2 = 3.56, df = 1, p = .06). There was no such difference
in opioid presence in the non-BD group (45.2 vs. 39.8%;
χ2 = 1.26, df = 1, p = .26).
The presence of antidepressants and benzodiazepines
were not significantly different between BD and non-BD
groups, with only higher frequency of mood stabilizer
(χ2 = 79.98, df = 1, p = < .0001) and antipsychotic use
(χ2 = 16.30, df = 1, p = < .0001) among BD decedents
differentiating the groups. There was a mean total of
Any Mood Stabilizer
An psycho c
Tricyclic An depressant
sub Any An depressant
Over the Counter Medica on
Other Prescrip on
Bipolar Disorder (n = 71)
Non-Bipolar Disorder (n = 498)
4.6 (SD = 1.9, range 0–10) different substances present
at time of death in the BD self-poisoning group,
significantly higher than the mean total of 3.9 (SD = 2.1, range
0–12) in the non-BD self-poisoning group (t = 2.78,
df = 567, p = .006).
Antidepressants, mood stabilizers, and antipsychotics
present in BD and non‑BD self‑poisoning groups
Figure 3 examines more closely the presence of
antidepressants with or without concomitant mood stabilizers
or antipsychotics. No antidepressant, mood stabilizer, or
antipsychotic was present in 15.5% of the total BD group.
Antidepressants were present in 58% of the BD group, of
which 39% did not have at least one mood stabilizer or
antipsychotic concomitantly present. As expected,
differences in medications present between BD and
nonBD groups were evident, with only the combination of
antidepressant plus antipsychotic not being significantly
Substances/medications present in BD and unipolar
depression self‑poisoning groups
A significantly higher proportion of
antipsychotics (45.1 vs. 22.4%, χ2 (1) = 15.63, p < .001) and mood
stabilizers (31.0 vs. 2.8%, χ2 (1) = 67.15, p < .001) was
noted among those with BD who died by self-poisoning
relative to those with unipolar depression who died by
This large study of suicide deaths in BD found that
selfpoisoning accounted for 36.7% of all deaths, which is
somewhat higher than other large studies (Chen et al.
2009; Hunt et al. 2006; Ösby et al. 2001), but within the
published range from all reports (Schaffer et al. 2015b).
Self-poisoning as the method of suicide for those with
BD was significantly correlated with female sex, prior
suicide attempts, and comorbid substance abuse. As far
as we are aware, this is a new finding that requires
replication from other sources, and if consistently reported
would be an important aspect of understanding which
subgroups within BD may be at particular risk of dying
by self-poisoning as compared to other methods.
Within the BD group, females accounted for 61.8% of
all self-poisoning suicides as compared to 32.8% by other
methods, a significant difference on univariate
analysis that was maintained in the regression model with an
OR 3.79. These findings have relevance to suicide risk
Any An depressant + Mood Stabilizer
Any An depressant + An psycho c
Any Mood Stabilizer or An psycho c
without An depressant
No An depressant, Mood Stabilizer or
An psycho c
Bipolar Disorder (n = 71)
Non-Bipolar Disorder (n = 498)
0 10 20 30 40 50
Fig. 3 Antidepressant, mood stabilizer and antipsychotic medications present at self-poisoning suicide among people with or without bipolar
disorder. *p ≤ .0001, **p = .007, ***p = .001
assessments which almost universally emphasize higher
risk for suicide in men, but which should also consider
that within the context of self-poisoning among people
with BD, women account for more suicides than men,
and self-poisoning is more predominantly used among
women with BD.
We found that opioids were the most common type
of identified lethal substance in both the BD and
nonBD groups. While the literature is clear that opioids are
a common source of lethal self-poisoning, we were
surprised for this to also be the case in people with BD. This
may be influenced by our finding of a non-significant
trend towards more frequent opioid lethality among
BD decedents with comorbid substance abuse, which
was not the case in the non-BD group. How the opioids
were obtained was not known to us, so we cannot
speculate on the proportion of prescribed or non-prescribed
sources, but growing evidence points to extensive access
to prescribed opioids among the population in general
(Tetrault and Butner 2015), including those at risk of
suicide (Ekholm et al. 2014; Madadi et al. 2013; Madadi
and Persaud 2014; West et al. 2015). Patients with BD are
known to have higher than expected rates of chronic pain
(Stubbs et al. 2015), and pain can certainly precipitate
suicidal behavior; however, the possibility of drug misuse
in the context of a substance use disorder or management
of psychological pain is clearly present. The significant
lethality associated with opioid self-poisoning further
increases the risk of opioid use in those at elevated risk
of suicide and suggests extra vigilance is required for this
population to ensure that appropriate means restriction
measures are in place.
Benzodiazepines and antidepressants were the next
most common classes of lethal substances used among
people with BD and were also the most common classes
present in the body at time of death. This indirectly
suggests that frequency of use was a larger factor than
specific lethality risk of the compounds. Coroner data did
not permit an examination of details of pharmacotherapy
such as indications for use, duration of therapy, or dosing
and prescribing patterns of medications, and therefore, it
is difficult to evaluate the appropriateness of
pharmacotherapy. Nonetheless, the greater frequency
benzodiazepine and antidepressant use as compared to lithium or
anticonvulsants suggests that a large number of patients
were not receiving BD guideline-concordant care at the
time of death. Furthermore, we found that 39% of those
with BD who had an antidepressant present at the time
of death did not have any concomitant mood stabilizer
or antipsychotic present. Nearly 1/6th of BD decedents
did not have any antidepressant, mood stabilizer, or
atypical antipsychotic present. To what degree these
findings relate to issues of adherence, physician selection of
therapy, or other factors is not known. The lack of a living
control group negates any possible discussion of risk of
suicide, however identifying what appears to be low rates
of guideline-concordant care (Goodwin and Consensus
Group of the British Association for 2009; Grunze et al.
2010; Yatham et al. 2013) is very concerning.
We can speculate that the relationship between
antidepressant/benzodiazepine use and suicide in BD may be
partially mediated through high rates of comorbid
anxiety (Schaffer et al. 2007; Schaffer et al. 2012), treatment
resistance (Yatham et al. 2005), illness severity (Marangell
et al. 2008), insomnia (Chung et al. 2015; Pompili et al.
2013b) and possible induction of mixed states or rapid
cycling (Pacchiarotti et al. 2013; Viktorin et al. 2014). The
findings may also be a function of higher prescriptions
rates for antidepressants and benzodiazepines among
women with BD compared to men with BD (Schaffer
et al. 2006; Weinstock et al. 2014). At a minimum,
clinicians and researchers should be aware of the pattern
of pharmacotherapy use at time of self-poisoning
suicide death, which deserves further clinical and research
Most of the extant literature on BD pharmacotherapy
and suicide has focused on lithium and
anticonvulsants (Cipriani et al. 2013; Oquendo et al. 2011;
Schaffer et al. 2015b). We found that only 7% of people with
BD who died by suicide had any lithium present at the
time of death, and the number of those with lethal
levels was below the threshold for reporting based on
privacy restrictions (i.e., n = 5). In contrast, anticonvulsants
were present at any level in 21% of BD suicide deaths,
most commonly being carbamazepine (14%), with the
other anticonvulsants below the privacy threshold. The
high number of deaths in which carbamazepine was
present was an unexpected finding, especially given the
declining use in clinical practice for BD, which is
currently estimated at 3% of BD patients (Mauer et al. 2014).
There is a modest literature suggesting higher rates of
suicide attempts during treatment with carbamazepine
monotherapy compared to lithium monotherapy
(Cipriani et al. 2013; Goodwin et al. 2003), but other factors
such as relative toxicity in overdose (Spiller et al. 1990)
and profile of patients who are prescribed carbamazepine
(Leon et al. 2012) may also be at play.
Presumably reflecting the changing practices in
management of BD, antipsychotics were more often present
than traditional mood stabilizers, with quetiapine and
olanzapine being the two antipsychotics that surpassed
the privacy threshold. While our data could not examine
relative anti-suicide effects, it is noteworthy that there
is only very sparse data on the impact of antipsychotic
medications on suicide risk in patients with BD (Koek
et al. 2012; Yerevanian et al. 2007).
The significantly higher proportion of past suicide
attempts and psychiatry/ER visits in the prior week
among the BD self-poisoning group relative to
unipolar depression self-poisoning group is consistent with
previous epidemiologic studies showing more frequent
mental health care contacts among BD and
schizophrenia suicide decedents (Schaffer et al. 2016). With regard
to lethal and present medications, the patterns of
differences in proportion of medications between BD and
unipolar self-poisoning groups are essentially reflective
of more commonly prescribed medications among these
There are a number of limitations that should be
considered when interpreting the results of this paper. First,
the diagnosis of BD was based on findings from coroner
investigations that rely on multiple sources, but do not
include a structured diagnostic assessment, and
misdiagnosis is therefore likely to have occurred in at least some
cases, most likely as false negatives. However, the BD
suicide group accounted for 6.2% of all suicide deaths, which
is very much in keeping with prior studies on the
proportion of BD decedents in large suicide samples (Karch et al.
2006; Schaffer et al. 2014; Takizawa 2012), and there was
10-fold greater use of mood stabilizers in the BD group
and small percentage (3%) in the non-BD group. Second,
the lack of a living control group limits any conclusions
related to risk impact of the pharmacotherapies
identified or of correlates of method type, as potential
causality could not be examined. Third, the comparison group
of non-BD suicide decedents is in fact a mix of other
diagnoses or no diagnosis at all. We chose to include a
broad comparison group in order to allow for basic
comparisons to be made for better characterization of the BD
subgroup. Furthermore, while we identified differences
between BD and unipolar depression groups, uncertainty
regarding diagnostic precision may limit the
interpretation of results. Fourth, the data on correlates of
self-poisoning as the method of suicide death were limited by the
variables available in coroner data, which did not include
many illness related factors that are known to influence
suicide risk (Schaffer et al. 2015a), and could not
differentiate medications taken as part of treatment as opposed
to only during the act of self-poisoning. Furthermore, it is
difficult to know whether the clinical differences impact a
greater propensity to using self-poisoning as the method
of attempt, greater access to large amounts of substances,
higher lethality as a result of underlying medical
vulnerability, or whether clinical aspects of the diagnosis were a
key aspect of the suicide death. Fifth, coroner’s toxicology
results necessarily examine only substances present at
the moment of death, and do not include information on
doses of prescribed medications or provide information
as to the source or intended use of the compound. This
is especially relevant for medication classes such as
opioids, which have broad-spectrum analgesic effects, but
are also subject to abuse and dependence. Furthermore,
this study could not answer important questions about
recent treatment such as how many people discontinued
a mood stabilizer days or weeks prior to their deaths.
This study highlights the specific characteristics of
people with BD who die by self-poisoning, a subgroup that
accounts for over a third of all BD suicide deaths. Female
sex, past suicide attempts, and comorbid substance abuse
each significantly increased the likelihood of
self-poisoning being the method of suicide among people with
bipolar disorder (BD), and clinicians should maintain a
high index of suspicion for self-poisoning suicide in these
Additional new findings related to the types of
substances present and lethal at the time of death are
informative at both a clinical and policy level, and if
replicated in other populations suggest the need for
heightened attention to issues of medication access and choices
for those at greater risk of suicide.
AS, LW, MS, CR, BG, LY, AL have each made substantial contributions to the
conception and design, or acquisition of data, or analysis and interpretation
of data, have been involved in drafting the manuscript or revising it critically
for important intellectual content, and have given final approval of the
manuscript. All authors read and approved the final manuscript.
The authors thank Yasunori Nishikawa for editorial and technical assistance.
Partial findings presented at IASR/AFSP International Summit on Suicide
Research in New York, United States.
Dr. Schaffer has in the past received speakers’ bureau honoraria, advisory
panel funding, and/or research grants from American Foundation for Suicide
Prevention, Brenda Smith Bipolar Disorder Research Fund, Bristol-Myers
Squibb; Lundbeck Canada, Ontario Mental Health Foundation, Otsuka, and
Sunovion. Dr. Weinstock has no conflict to report. Dr. Sinyor has received grant
support from the American Foundation for Suicide Prevention, Physicians’
Services Incorporated (PSI) Foundation, the University of Toronto, Department
of Psychiatry Excellence Fund and the Brenda Smith Bipolar Disorder Research
Fund. Ms. Reis has no conflict to report. Dr. Goldstein has received research
support from Pfizer, is a consultant for Bristol-Myers Squibb, and has received
speakers honoraria from Purdue Pharma. Dr. Yatham has been a member of
advisory boards and/or received research grants and/or been a speaker for
Astra Zeneca, DSP, Janssen, Lilly, GSK, Bristol Myers Squibb, Lundbeck, Novartis,
Servier, Sunovion and Pfizer. Dr. Levitt has received research grants from
Janssen Ortho, AstraZeneca, Great West Life Insurance, and Eli Lilly Canada and has
acted as a consultant for Janssen Ortho.
Springer Nature remains neutral with regard to jurisdictional claims in
published maps and institutional affiliations.
Angst J , Angst F , Gerber-Werder R , Gamma A. Suicide in 406 mood-disorder patients with and without long-term medication: a 40 to 44 years' followup . Arch Suicide Res . 2005 ; 9 ( 3 ): 279 - 300 .
Camidge DR , Wood RJ , Bateman DN . The epidemiology of self-poisoning in the UK . Br J Clin Pharmacol . 2003 ; 56 ( 6 ): 613 - 9 .
Chen YY , Lee MB , Chang CM , Liao SC . Methods of suicide in different psychiatric diagnostic groups . J Affect Disord . 2009 ; 118 ( 1 ): 196 - 200 .
Chung KH , Li CY , Kuo SY , Sithole T , Liu WW , Chung MH . Risk of psychiatric disorders in patients with chronic insomnia and sedative-hypnotic prescription . J Clin Sleep Med . 2015 ; 11 ( 5 ): 543 - 51 .
Cipriani A , Hawton K , Stockton S , Geddes JR . Lithium in the prevention of suicide in mood disorders: updated systematic review and meta-analysis . BMJ . 2013 ; 346 :f3646.
Clements C , Morriss R , Jones S , Peters S , Roberts C , Kapur N. Suicide in bipolar disorder in a national English sample , 1996 - 2009 : frequency, trends and characteristics. Psychol Med . 2013 ; 43 ( 12 ): 2593 - 602 .
Crump C , Sundquist K , Winkleby MA , Sundquist J. Comorbidities and mortality in bipolar disorder: a Swedish national cohort study . JAMA Psychiatry . 2013 ; 70 ( 9 ): 931 - 9 .
Dennehy EB , Marangell LB , Allen MH , Chessick C , Wisniewski SR , Thase ME . Suicide and suicide attempts in the systematic treatment enhancement program for bipolar disorder (STEP-BD) . J Affect Disord . 2011 ; 133 ( 3 ): 423 - 7 .
Ekholm O , Kurita GP , Højsted J , Juel K , Sjøgren P. Chronic pain, opioid prescriptions, and mortality in Denmark: a population-based cohort study . PAIN® . 2014 ; 155 ( 12 ): 2486 - 90 .
Goodwin FK , Fireman B , Simon GE , Hunkeler EM , Lee J , Revicki D. Suicide risk in bipolar disorder during treatment with lithium and divalproex . JAMA . 2003 ; 290 ( 11 ): 1467 - 73 .
Goodwin GO , Consensus Group of the British Association for Psychopharmacology. Evidence-based guidelines for treating bipolar disorder: revised second edition-recommendations from the British Association for Psychopharmacology . J Psychopharmacol . 2009 ; 23 ( 4 ): 346 - 88 .
Gos T , Günther K , Bielau H , Dobrowolny H , Mawrin C , Trübner K , et al. Suicide and depression in the quantitative analysis of glutamic acid decarboxylase-immunoreactive neuropil . J Affect Disord . 2009 ; 113 ( 1 ): 45 - 55 .
Grunze H , Vieta E , Goodwin GM , Bowden C , Licht RW , Möller HJ , Kasper S. The World Federation of Societies of Biological Psychiatry (WFSBP) guidelines for the biological treatment of bipolar disorders: update 2010 on the treatment of acute bipolar depression . World J Biol Psychiatry . 2010 ; 11 ( 2 ): 81 - 109 .
Harris EC , Barraclough B. Suicide as an outcome for mental disorders . A metaanalysis. Br J Psychiatry . 1997 ; 170 ( 3 ): 205 - 28 .
Høyer EH , Mortensen PB , Olesen AV . Mortality and causes of death in a total national sample of patients with affective disorders admitted for the first time between 1973 and 1993 . Br J Psychiatry . 2000 ; 176 ( 1 ): 76 - 82 .
Hunt IM , Kapur N , Robinson J , Shaw J , Flynn S , Bailey H , et al. Suicide within 12 months of mental health service contact in different age and diagnostic groups . Br J Psychiatry . 2006 ; 188 ( 2 ): 135 - 42 .
Ilgen MA , Bohnert AS , Ignacio RV , McCarthy JF , Valenstein MM , Kim HM , et al. Psychiatric diagnoses and risk of suicide in veterans . Arch Gen Psychiatry . 2010 ; 67 ( 11 ): 1152 - 8 .
Isomets ET , Henriksson M. Suicide in bipolar disorder in Finland . Am J Psychiatry . 1994 ; 151 ( 7 ): 169 - 77 .
Kapur N , Turnbull P , Hawton K , Simkin S , Sutton L , Mackway-Jones K , et al. Self-poisoning suicides in England: a multicentre study . QJM . 2005 ; 98 ( 8 ): 589 - 97 .
Karch DL , Barker L , Strine TW . Race/ethnicity, substance abuse, and mental illness among suicide victims in 13 US states: 2004 data from the National Violent Death Reporting System . Inj Prev . 2006 ; 12 (suppl 2): 22 - 7 .
Keks NA , Hill C , Sundram S , Graham A , Bellingham K , Dean B , et al. Evaluation of treatment in 35 cases of bipolar suicide . Aust NZ J Psychiatry . 2009 ; 43 ( 6 ): 503 - 8 .
Kessing LV , Søndergård L , Kvist K , Andersen PK . Suicide risk in patients treated with lithium . Arch Gen Psychiatry . 2005 ; 62 ( 8 ): 860 - 6 .
Koek RJ , Yerevanian BI , Mintz J. Subtypes of antipsychotics and suicidal behavior in bipolar disorder . J Affect Disord . 2012 ; 143 ( 1 ): 27 - 33 .
Leon AC , Solomon DA , Li C , Fiedorowicz JG , Coryell WH , Endicott J , et al. Antiepileptic drugs for bipolar disorder and the risk of suicidal behavior: a 30-year observational study . Am J Psychiatry . 2012 ; 169 ( 3 ): 285 - 91 .
Madadi P , Hildebrandt D , Lauwers AE , Koren G. Characteristics of opioid-users whose death was related to opioid-toxicity: a population-based study in Ontario , Canada. PLoS ONE . 2013 ; 8 ( 4 ): e60600 .
Madadi P , Persaud N. Suicide by means of opioid overdose in patients with chronic pain . Curr Pain Headache Rep . 2014 ; 18 ( 11 ): 1 - 4 .
Marangell LB , Dennehy EB , Wisniewski SR , Bauer MS , Miyahara S , Allen MH , et al. Case-control analyses of the impact of pharmacotherapy on prospectively observed suicide attempts and completed suicides in bipolar disorder: findings from STEP-BD . J Clin Psychiatry . 2008 ; 69 ( 6 ): 916 - 22 .
Mauer S , Alahmari R , Vöhringer PA , Vergne DE , Lövdahl H , Correa E , et al. International prescribing patterns for mood illness: the international mood network (IMN) . J Affect Disord . 2014 ; 1 ( 167 ): 136 - 9 .
Oquendo MA , Galfalvy HC , Currier D , Grunebaum MF , Sher L , Sullivan GM , et al. Treatment of suicide attempters with bipolar disorder: a randomized clinical trial comparing lithium and valproate in the prevention of suicidal behavior . Am J Psychiatry . 2011 ; 168 ( 10 ): 1050 - 6 .
Ösby U , Brandt L , Correia N , Ekbom A , Sparén P. Excess mortality in bipolar and unipolar disorder in Sweden . Arch Gen Psychiatry . 2001 ; 58 ( 9 ): 844 - 50 .
Pacchiarotti I , Bond DJ , Baldessarini RJ , Nolen WA , Grunze H , Licht RW , et al. The International Society for bipolar Disorders (ISBD) task force report on antidepressant use in bipolar disorders . Am J Psychiatry . 2013 ; 170 ( 11 ): 1249 - 62 .
Pompili M , Gonda X , Serafini G , Innamorati M , Sher L , Amore M , et al. Epidemiology of suicide in bipolar disorders: a systematic review of the literature . Bipolar Disord . 2013a; 15 ( 5 ): 457 - 90 .
Pompili M , Innamorati M , Forte A , Longo L , Mazzetta C , Erbuto D , et al. Insomnia as a predictor of high-lethality suicide attempts . Int J Clin Pract . 2013b ; 67 ( 12 ): 1311 - 6 .
Rihmer Z , Barsi J , Arató M , Demeter E. Suicide in subtypes of primary major depression . J Affect Disord . 1990 ; 18 ( 3 ): 221 - 5 .
Schaffer A , Cairney J , Cheung AH , Veldhuizen S , Levitt AJ . Use of treatment services and pharmacotherapy for bipolar disorder in a general populationbased mental health survey . J Clin Psychiatry . 2006 ; 67 ( 3 ): 386 - 93 .
Schaffer A , Cairney J , Veldhuizen S , Cheung A , Levitt A. Comparison of antidepressant use between subjects with bipolar disorder and major depressive disorder with or without comorbid anxiety . J Clin Psychiatry . 2007 ; 68 ( 11 ): 1785 - 92 .
Schaffer A , Isometsä ET , Tondo L , Moreno H , Turecki G , Reis C , et al. International Society for Bipolar Disorders Task Force on Suicide: meta-analyses and meta-regression of correlates of suicide attempts and suicide deaths in bipolar disorder . Bipolar Disord . 2015a; 17 ( 1 ): 1 - 6 .
Schaffer A , Isometsä ET , Tondo L , Moreno DH , Sinyor M , Kessing LV , et al. Epidemiology , neurobiology and pharmacological interventions related to suicide deaths and suicide attempts in bipolar disorder: part I of a report of the International Society for Bipolar Disorders Task Force on Suicide in Bipolar Disorder . Aust NZ J Psychiatry . 2015b ; 49 ( 9 ): 785 - 802 .
Schaffer A , McIntosh D , Goldstein BI , Rector NA , McIntyre RS , Beaulieu S , et al. The CANMAT task force recommendations for the management of patients with mood disorders and comorbid anxiety disorders . Ann Clin Psychiatry . 2012 ; 24 ( 1 ): 6 - 22 .
Schaffer A , Sinyor M , Reis C , Goldstein BI , Levitt AJ . Suicide in bipolar disorder: characteristics and subgroups . Bipolar Disord . 2014 ; 16 ( 7 ): 732 - 40 .
Schaffer A , Sinyor M , Kurdyak P , Vigod S , Sareen J , Reis C , et al. Populationbased analysis of health care contacts among suicide decedents: identifying opportunities for more targeted suicide prevention strategies . World Psychiatry . 2016 ; 15 ( 2 ): 135 - 45 .
Sinyor M , Howlett A , Cheung AH , Schaffer A. Substances used in completed suicide by overdose in Toronto: an observational study of coroner's data . Can J Psychiatry . 2012 ; 57 ( 3 ): 184 - 91 .
Sinyor M , Schaffer A , Remington G . Suicide in schizophrenia: an observational study of coroner records in Toronto . J Clin Psychiatry . 2015 ; 76 ( 1 ): 98 - 103 .
Sinyor M , Schaffer A , Streiner DL . Characterizing suicide in Toronto: an observational study and cluster analysis . Can J Psychiatry . 2014 ; 59 ( 1 ): 26 - 33 .
Spiller HA , Krenelok EP , Cookson E. Carbamazepine overdose: a prospective study of serum levels and toxicity . J Toxicol Clin Toxicol . 1990 ; 28 ( 4 ): 445 - 58 .
Stubbs B , Eggermont L , Mitchell AJ , De Hert M , Correll CU , Soundy A , et al. The prevalence of pain in bipolar disorder: a systematic review and largescale meta-analysis . Acta Psychiatr Scand . 2015 ; 131 ( 2 ): 75 - 88 .
Takizawa T. Suicide due to mental diseases based on the Vital Statistics Survey Death Form]. [Nihon koshu eisei zasshi] . Jpn J Public Health . 2012 ; 59 ( 6 ): 399 - 406 .
Tetrault JM , Butner JL . Focus: addiction: non-medical prescription opioid use and prescription opioid use disorder: a review . Yale J Biol Med . 2015 ; 88 ( 3 ): 227 .
Viktorin A , Lichtenstein P , Thase ME , Larsson H , Lundholm C , Magnusson PK , et al. The risk of switch to mania in patients with bipolar disorder during treatment with an antidepressant alone and in combination with a mood stabilizer . Am J Psychiatry . 2014 ; 171 ( 10 ): 1067 - 73 .
Weinstock LM , Gaudiano BA , Epstein-Lubow G , Tezanos K , Celis-deHoyos CE , Miller IW . Medication burden in bipolar disorder: a chart review of patients at psychiatric hospital admission . Psychiatry Res . 2014 ; 216 ( 1 ): 24 - 30 .
West NA , Severtson SG , Green JL , Dart RC . Trends in abuse and misuse of prescription opioids among older adults . Drug Alcohol Depend . 2015 ; 1 ( 149 ): 117 - 21 .
Yatham LN , Kennedy SH , O'Donovan C , Parikh S , MacQueen G , McIntyre R , et al. Canadian Network for Mood and Anxiety Treatments (CANMAT) guidelines for the management of patients with bipolar disorder: consensus and controversies . Bipolar Disord . 2005 ; 7 (s3): 5 - 69 .
Yatham LN , Kennedy SH , Parikh SV , Schaffer A , Beaulieu S , Alda M , et al. Canadian Network for Mood and Anxiety Treatments (CANMAT) and International Society for Bipolar Disorders (ISBD) collaborative update of CANMAT guidelines for the management of patients with bipolar disorder: update 2013 . Bipolar Disord . 2013 ; 15 ( 1 ): 1 - 44 .
Yerevanian BI , Koek RJ , Mintz J. Bipolar pharmacotherapy and suicidal behavior: part 3: impact of antipsychotics . J Affect Disord . 2007 ; 103 ( 1 ): 23 - 8 .