Completed suicides: Is there method in their madness? correlates of choice of method for suicide in an Australian sample of suicides

Abstract

Suicides choose different methods for their act, and it is of interest to ask whether those choosing different methods have different characteristics. A complete sample of 1051 suicides from a single region of Australia (Tasmania) was studied for differences in their characteristics by their method for suicide. Clear differences emerged between those using different methods. Those using gas inhalation were reacting to interpersonal conflict or loss and misused alcohol. Those using jumping had ongoing psychiatric problems and a history of suicidal behavior. Those using firearms were characterized by anger and were more likely to commit murder-suicide than those using other methods. These differences were explored for their implications for prevention.

Full text

133
Clinical Neuropsychiatry (2010) 7, 4/5,
© 2010 Giovanni Fioriti Editore s.r.l.
133-140
SUBMITTED AUGUST 2010, ACCEPTED OCTOBER 2010
COMPLETED SUICIDES: IS THERE METHOD IN THEIR MADNESS? CORRELATES OF CHOICE OF
METHOD FOR SUICIDE IN AN AUSTRALIAN SAMPLE OF SUICIDES
Janet Haines, Christopher L. Williams, & David Lester
Abstract
Suicides choose different methods for their act, and it is of interest to ask whether those choosing different
methods have different characteristics. A complete sample of 1051 suicides from a single region of Australia (Tasmania)
was studied for differences in their characteristics by their method for suicide. Clear differences emerged between
those using different methods. Those using gas inhalation were reacting to interpersonal conflict or loss and misused
alcohol. Those using jumping had ongoing psychiatric problems and a history of suicidal behavior. Those using
firearms were characterized by anger and were more likely to commit murder-suicide than those using other methods.
These differences were explored for their implications for prevention.
Key Words: suicide, Australia, methods for suicide
Declaration of interest: none
Janet Haines, Christopher L. Williams, University of Tasmania
David Lester, The Richard Stockton College of New Jersey
Corresponding author
David Lester
Psychology Program
The Richard Stockton College of New Jersey
Pomona, NJ 08240
lesterd@stockton.edu
No psychological differences have been found to
differentiate those using different methods for suicide,
and Lester (1997) concluded that situational variables
(such as the availability of methods for suicide) might
be more important in determining the choice of method
for suicide than psychological variables. However,
some personal characteristics of suicides using different
methods have been identified. In the United Kingdom,
Indians and Pakistanis prefer hanging more than other
ethnic groups, the elderly use drowning more than
younger adults, and those with prior psychiatric
disorders more often use drugs and drowning (Scott
1994). In New South Wales (Australia), Burnley (1995)
found that youths, males, rural inhabitants and farmers
and transport workers used firearms more often for
suicide than do other groups.
Hitherto, much of the research on the method
chosen for suicide has focused on attempted suicides,
but there is good evidence that there are sufficient
differences between attempted and completed suicides
to warrant separate consideration of completed suicide
(Beautrais 2001a). Much of the research of choice of
method by completed suicides has focused on a single
method, such as self-poisoning (Reith et al. 2003), on a
limited number of methods, such as firearms versus
jumping (De Moore & Robertson 1999, De Leo et al.
When asked about their preferred method if they
were going to complete suicide, most people can choose
one method, and many of these state that they would
not change methods if their preferred method was
unavailable. This phenomenon has led suicidologists
to suggest restricting access to methods for suicide
(such as gun control and fencing in popular suicide
venues) as a tactic for preventing suicide (e.g., Clarke
& Lester 1989).
The methods for suicide are viewed quite
differently. Firearms are viewed as a quick, painful,
irreversible, dramatic, masculine and messy method
whereas an overdose of pills is viewed as slow, painless,
easy, cowardly, feminine, tidy and planned (Lester
1987). Choice of method for suicide should, therefore,
have associations with characteristics of the suicidal
person. For example, if suicide by firearm is seen as
masculine whereas suicide by pills is seen as feminine,
then it might be predicted that men would choose
firearms for suicide more than women (and perhaps
that masculine men and women should choose firearms
for suicide more than feminine men and women).
Epidemiological studies confirm that men do chose
firearms for suicide more than do women (Lester 1993),
but no studies have been published on masculinity/
feminity and choice of method for suicide.

134
Janet Haines et al.
Clinical Neuropsychiatry (2010) 7, 4/5
2002), on restricted samples, such as male farmers
(Hawton et al. 1998) or the elderly (Salib & Maxious
2002).
The aim of the present study was to undertake a
comprehensive examination of the factors that
differentiate those using different methods for suicide
using a sample of all the suicides in a given region in a
given time period.
Method
The Sample
This research studied 1,051 completed suicides
from a twenty-year period in Tasmania. Information
about the cases was obtained from the coroners inquest
files held at the Tasmanian Archives Office and the
Tasmanian Department of Justice. All the deaths had
been identified by the coroner as being caused by
suicide and consisted of all suicides in that period. The
method of suicide was missing in eight cases, and so
the sample was reduced to 1,043 cases. Tasmanian
coroners are appointed through the Magistrates Court,
and the scope of their powers in described in the website
www.magistratescourt.tas.gov.au/divisions/coronial.
How the decisions of Tasmanian coroners compare to
those made by coroners and medical examiners in other
jurisdictions has never been studied.
The Variables
A possible list of variables was generated and
placed on a record sheet. Then 50 files of suicidal deaths
were examined which resulted in suggestions for new
variables. These were then added to the original record
sheet, and another 50 files of suicidal deaths were
examined, and further adjustments made to the record
sheet. This new record sheet was then applied to the
complete sample of 1043 suicides.
Data for the following variables were collected:
marital status, age, sex, living arrangements, employ-
ment status, suicide history and stated intention, medical
and psychiatric history, psychological state prior to the
act, psychological symptoms in the days and weeks
leading up to the act, reasons for the suicide and method
of suicide. Two raters coded the data and their inter-
rater agreement was 97%.
Results
The number of deaths from each method is shown
in table 1. The use of firearms was the most common
method (40%), followed by medications (18%), gas
(13%) and hanging (11%).
In table 2, those using each method are compared
for each of the variables coded. It can be seen that the
majority of comparisons were statistically significant.
(1) Men were more likely to use gas and firearms than
women. (2) Single individuals were more likely to use
violent methods (firearms, hanging and jumping),
married individuals poisons, divorced individuals gas
and widowed individuals drowning. (3) Those
employed were more likely to use gas. (4) Differences
by living arrangements were similar to those by marital
status (for example, those living with a spouse used
poisons more often). (5) Alcohol was ingested around
the time of the suicide most by those using gas, firearms
and medications. (6) A history of attempted suicide was
most common in those using hanging, jumping and
medications. (7) A suicide note was left more often by
those using gas, guns and medications. (8) Those using
medication, poisons and drowning had the poorest
health. (9) Psychiatric problems were most common in
those using jumping. (10) Sadness in the days preceding
the suicide was most common in those using hanging,
drowning and poison, while anger was more common
in those guns and poisons. (11) The reasons for suicide
varied by method in line with some of the above
differences, with conflict motivating more suicides by
gun, psychiatric disturbance suicides by jumping, and
physical problems those using medications, poison and
drowning.
Two problems with these simple two-by-two tables
are that many of the variables are inter-correlated and
there seems to be an impact of age. For example,
widows used drowning and poisons more often, but this
may be a function of the fact that widows are generally
older than single, married and divorced individuals. In
order to control for these problems, a number of the
dichotomous variables that differed significantly by the
method used for the suicide, along with age, were
subjected to factor analysis using a principal
components extraction and a varimax rotation. The
variables included were those with few missing data
points. The results are shown in table 3. Eight factors
(with eigenvalues greater than one) were extracted, and
the high loadings are shown. (The choice of loadings
greater than 0.38 included every variable in the
solution.) The factors are labelled based on the high
loadings and appear to represent: (1) alcohol use, (2)
depressed, (3) medically ill, (4) mentally disturbed, (5)
angry, (6) female and extrapunitive, (7) anxious, and
(8) interpersonal conflict and loss.
Those choosing the different methods were then
compared on these eight factor scores and the means
scores differing significantly from zero are shown in
table 4. (Factor scores have a mean of zero).
Discussion
The aim of this study was to examine the personal
and risk factors associated with the choice of a method
for suicide. The results indicated that those using the
possible methods for suicide differed on most of the
variables studied.
Based on the results portrayed in tables 2 and 4,
people who selected gas inhalation as a method for
suicide were likely to have experienced an interpersonal
conflict or loss that lead to distress and threats of suicide
in the time leading up to the death. There would have
been an increasing use of alcohol, including
consumption immediately prior to death. The use of
alcohol would result in disinhibition, resulting in a
suicidal action. There was little evidence of chronic
medical, psychiatric or psychological problems prior
to death. Despite the interpersonal trigger, there was

135
Methods for suicide
Clinical Neuropsychiatry (2010) 7, 4/5
Table 1. The number of suicides by each method (n = 1043) and their age
Frequency Percentage Mean age SD
Gas inhalation 131 12.6 40.2 14.6
Firearm 419 40.2 40.1 18.1
Hanging 118 11.3 46.5 18.4
Jumping from a height 27 2.6 40.8 16.0
Drowning 83 8.0 56.9 16.6
Self-poisoning (pills) 193 18.5 43.0 15.1
Self-poisoning (poison) 31 3.0 48.8 19.5
Other 41 3.9 51.9 17.3
For age, F = 12.90, df = 7,1025, p < .001
Table 2. Percent of persons who used a particular method of suicide who also exhibited the identified characteristics
(N=1043)


9DULDEOH /HYHO *DV *XQ +DQJ -XPS 'URZQ 3LOOV 3RLVRQ 2WKHU ;ñ 1 GI
'HPRJUDSKLF 
6H[ 0DOH           
0DULWDO 6LQJOH           
6WDWXV 0DUULHG           
 6HSGLY           
 :LGRZHG           
(PSOR\PHQW <HV           
/LYLQJ 6SRXVH           
$UUDQJHPHQWV )DPLO\           
 2WKHUV           
 $ORQH           
7R[LFRORJ\ 
+\SQRWLF <HV           
%DUELWXUDWH <HV           
0LQRUWUDQT <HV           
0DMRUWUDQT <HV           
$QWLGHSUHVV <HV           
$QDOJHVLFV <HV           
2WKHUGUXJV <HV           
2WKHUSRLVRQ <HV           
$OFRKRO <HV           

136
Janet Haines et al.
Clinical Neuropsychiatry (2010) 7, 4/5
Table 2. Continued

9DULDEOH /HYHO *DV *XQ +DQJ -XPS 'URZQ 3LOOV 3RLVRQ 2WKHU ;ð 1 GI
6XLFLGDO
KLVWRU\
  
3UHYDWWHPSW <HV           
6XLF0RGHO <HV           
6WDWHGLQWHQW <HV           
6XLFLGHQRWH <HV           
0HGLFDODQG
SV\FKLDWULF
KLVWRU\
  
'RF&RQVXOW <HV           
0HG6XS <HV           
3RRUKHDOWK <HV           
*HQKRVSLWDO <HV           
3V\&RQVXOW <HV           
3V\6XS <HV           
3V\KRVSLWDO <HV           
3V\FKRORJLFDO
VWDWHV
SUHFHGLQJ
GHDWK
  
1RUPDOVHOI <HV           
8SVHW <HV           
'LVWUHVVHG <HV           
$QJU\ <HV           
6DG <HV           
/RQHO\ <HV           
(UUDWLF <HV           
,PSXOVLYH <HV           
&RQIXVHG <HV           
&KHHUIXO <HV           
'UXQN <HV           
/DELOHPRRG <HV           
&DOP <HV           

no particular wish to punish others (extrapunitive
motives). An inability to cope without the relationship
and the partner, rather than angry or punitive feelings,
would have been evident. These themes may have been
apparent in the content of the suicide notes that were
frequently left by these individuals. The interpersonal
conflict, in combination with increased alcohol use, and
in the absence of any other specific predictors, seems
to be sufficient to result in suicidal behavior in those
using gas as a method. Reducing or removing access to
the means may enhance the chances that the passage of
time and support interventions would resolve these
transient difficulties for the suicidal individual.
The risk factors for selecting a firearm as a method
of suicide were not dissimilar to those associated with
gas inhalation although interpersonal conflict or loss
was not a strong motivation in these cases. The nature
of the interpersonal difficulties for individuals who used
a firearm was different and was characterized by anger.
The motivation was not extrapunitive in the normal
sense of wishing to punish someone in terms of making
them psychologically suffer. An inability to cope with
anger or a desire to kill both themselves and the
perceived offending party may have been more
apparent. This view is supported by the inclusion of
homicide-suicide as a characteristic of this group. The
disinhibiting effects of alcohol, when combined with
an angry response, would have lead to a violent
outcome. A combination of anger and alcohol could
also have resulted in increased impulsiveness which
would explain the lack of stated suicidal intent in the
time leading up to death. If reactive aggression triggers
violence that manifests as suicide and, in some cases,
homicide as well, it must be speculated whether

137
Methods for suicide
Clinical Neuropsychiatry (2010) 7, 4/5

9DULDEOH /HYHO *DV *XQ +DQJ -XPS 'URZQ 3LOOV 3RLVRQ 2WKHU ;ð 1 GI
5HDVRQIRU
VXLFLGH
  
3V\'LVWXUE <HV           
3K\V,OOQHVV <HV           
&RQIOLFW <HV           
,VRODWLRQ <HV           
/LIHHYHQW <HV           
+RP±
VXLFLGH
<HV           
([WUDSXQLWLYH <HV           
5HPRUVH <HV           
0RGHOOLQJ <HV           
3V\FKRORJLFDO
V\PSWRPV
  
$JLWDWLRQ <HV           
6OHHSGLVWXUE <HV           
:LWKGUDZDO <HV           
'HSUHVVLRQ <HV           
$Q[LHW\ <HV           
+\SRFKRQG <HV           
6XLUXPLQDW <HV           
3V\FKRVLV <HV           
$OFDEXVH <HV           
S
S
  

Table 2. Continued
managing access to firearms would increase the
likelihood that the angry response is handled in some
other way such as assault or property damage. These
outcomes, although not desirable, are clearly preferable
to behaviours that result in the death of one or more
individuals. When premeditation is absent, any
intervention that extends the time between desire for
violence and action, such as making firearms less
readily available, could potentially decrease the
likelihood of a tragic outcome.
There were similarities in the characteristics
associated with hanging and jumping from heights as
methods of suicide. People who selected hanging as a
method of suicide were characterized by having
ongoing psychiatric concerns, with anxiety and
depression being particularly problematic. There was
a heavy use of psychiatric services. The ongoing
psychiatric disturbance was the trigger for the suicidal
behavior. These individuals had previously attempted
suicide. There was little evidence of alcohol being a
factor in these cases, and so the disinhibiting effects of
alcohol were absent. Further, there was little evidence
of anger or extrapunitive intention, and interpersonal
conflict and loss were not important factors. There was
little indication that there were ongoing medical
concerns. Statements of suicidal intention were absent,
and notes rarely left. There was little evidence of distress
leading up to death. The suicidal action appears to have
been a more thought-out response to chronic depression.
In terms of preventing suicide, it is worthy of note that,
among this group, there was no apparent escalation of
distress or increased talk of suicide for these chronically
depressed and anxious people in the time leading up to
death. It would appear that depression and other
common psychiatric symptoms, without additional risk
factors being evident, could result in death by hanging.
With the access to ongoing psychiatric support available
to these individuals, mental health professionals should
routinely monitor both suicidal ideation and, because
hanging is a solitary activity, should endeavor to
increase the availability of support and companionship
outside of the therapeutic relationship.
The people who chose to jump from heights as a
method of suicide were similar in many ways to those
who selected hanging. In particular, they experienced
chronic anxiety and were users of psychiatric services.
In addition, their psychiatric disturbance was identified
as a motive for the suicide. However, in contrast to the
people who hanged themselves, those who jumped from
heights were not depressed and did not have a history
of problems with depression.
Drowning, self-poisoning with medication, and

138
Janet Haines et al.
Clinical Neuropsychiatry (2010) 7, 4/5
1 2 3 4 5 6 7 8
age 75
sex 61
previous attempt 39
note 73
distressed 57
angry 69
sad 84
erratic 66
drunk 94
psychologically disturbed 49 50
physical illness 70
conflict 43 40
stressful life events -47 -47
murder-suicide 79
extrapunitive 64
sleep 71
depression 82
anxiety 75
psychosis 72
alcohol abuse 94
% variance 14.7 8.9 7.7 7.0 6.5 5.7 5.4 5.2
hang guns drown jump pills gas poison other
1 2 3 4 5 6 7 8
Factor score
1 Alcohol use -34 13 -22
2 Depressed -16 37
3 Medically ill 49 -20
4 Mentally disturbed -16 66
5 Angry 18 -20
6 Female -27 64 -17
7 Anxious -15 42 28 -19
8 Conflict -24 -16 32
Table 3. Factor Analysis (n = 998)
(Loadings > 0.38 are shown, with decimal points omitted)
Table 4. Mean factors scores significantly different from zero by method
(decimal points omitted)

139
Methods for suicide
Clinical Neuropsychiatry (2010) 7, 4/5
self-poisoning with non-ingestant substances shared
some similarities. For the individuals who selected
drowning as a method of suicide, medical and
psychiatric problems were evident. In particular, these
individuals were in poor health and were users of
medical services. Indeed, physical illness was identified
as a motive for the suicidal action. Anxiety and
depression were experienced by these individuals and
assistance was sought from medical practitioners.
However, these individuals also wanted to punish others
by their actions, although there were not any particularly
stressful events that precipitated the suicide. Although
their actions were extrapunitive, they were not
particularly angry leading up to death. Therefore, the
extrapunitive motivation, in the absence of angry
feelings or specified interpersonal conflict, would
indicate that, in general, intimate relationships were
poor and contributed to the decision to suicide.
It is interesting that an extrapunitive influence was
also evident in this group. It has been postulated that,
when physically ill individuals experience personal
criticism, even minor criticism, by those providing social
support, they view themselves as being a burden, and
there is a resultant increase in suicidal ideation (Brown
& Vinokur 2003). It could also be argued that this process
creates feelings of ill-will in the suicidal individual,
meaning that the person will use their suicidal behavior
as a means of punishing the significant other for the
criticism or perceived withdrawal of support.
Individuals who chose to ingest an overdose of
medication shared the medical and extrapunitive
characteristics of those who chose drowning. In
addition, they were anxious like the individuals who
chose drowning but who were not depressed. These
individuals had health problems resulting in the use of
medical services. Physical illness was a contributing
factor to their suicides. Punishing others was a risk
factor for overdosing. No particularly stressful life event
triggered the suicidal behavior. Ongoing anxiety was a
problem that resulted in the use of psychiatric services.
There was little evidence of anger leading up to death
despite the extrapunitive motivation, and there was little
evidence of interpersonal conflict as a motive for
suicide. Again, it would be more likely that the suicidal
behavior was a function of the poor quality of
interpersonal relationships rather than a reaction to a
specific interpersonal event. Stated intention to suicide
was not strongly related to this method.
It has been proposed that motivations for attempted
suicide may be divided into two broad categories:
internal perturbation and extrapunitive/manipulative
(e.g., Holden & Kroner 2003, Holden et al. 1998, Johns
& Holden 1997). It would appear from this exami-
nation of completed suicide that the extrapunitive/
manipulative group of motivations do not fit well
together. For example, it is evident that people who
take an overdose may do so for extrapunitive reasons,
but they are not particularly angry and have not
experienced a specific interpersonal conflict as a
trigger. Extrapunitive motives may relate to a wish to
make the person suffer for a perceived injustice. Angry
feelings seem to relate to outrage at the actions of
another without the desire to punish, and interpersonal
conflict precipitates feelings associated with an
inability or lack of desire to cope without the rela-
tionship or the partner. These three factors seem to be
related separately to different methods of suicide.
The medical and extrapunitive factors associated
with drowning and overdosing were also present for
the selection of self-poisoning with non-ingestant
substances as a method of suicide. No other factors were
associated with this method. The effect of perceived
support withdrawal in a physically unwell individual
(Brown & Vinokur 2003) is probably most evident for
those ingesting poison, without the confounding effect
of extraneous variables influencing the decision to sui-
cide by this method.
Finally, the other or miscellaneous group of
methods were made up of low frequency and often
bizarre methods of suicide. These individuals had a
history of psychiatric disturbance and were users of
mental health services. Their psychiatric disturbance
was the trigger for the suicidal action and resulted in
erratic behavior leading up to death. Psychotic and
psychotic-like symptoms were experienced by these
individuals and the intensity of these symptoms
probably increased in the time leading up to death. There
was a history of previous suicide attempts that were a
reflection of their ongoing psychiatric concerns. There
was little evidence of an extrapunitive motivation for
suicide and little evidence of interpersonal conflict or
overt expression of suicidal intentions.
Although there was some evidence of an
exacerbation of symptoms in the time leading up to
death, there appeared to be no other factor, such as a
stressful life event or an interpersonal conflict, that
triggered the suicidal action. Stressful life events and
social factors have been demonstrated by others to have
little impact on the suicides of people with serious
mental illness (Cooper et al. 2002, King et al. 2001).
This is important information in relation to the
evaluation of risk of suicide among individuals
experiencing psychotic symptoms.
In conclusion, the present study, based on a large
sample of completed suicides, has identified different
patterns leading up to the suicide in those choosing
different methods of suicide. These patterns differed in
characteristics such as planning, the presence of
psychiatric disorders, and the role of precipitating
events. These differences have implications for planning
treatment and for possibilities for prevention.
Many of the factors identified as contributing to
the suicides have been identified in previous research
on suicide. For example, the role of depressive
symptoms is well-documented (e.g., Snowdon &
Baume 2002) as is anxiety (e.g., Busch et al. 1998).
The present results indicate that these factors may not
apply to all suicides, but only to specific sub-groups,
in the present case, categorized by preferred methods
for suicide.
References
Beautrais AL (2001a). Suicides and serious suicide attempts:
Two populations or one? Psychological Medicine 31, 837-
845.
Brown SL, Vinokur AD (2003). The interplay among risk factors
for suicidal ideation and suicide. American Journal of
Community Psychology 35, 557-562.

140
Janet Haines et al.
Clinical Neuropsychiatry (2010) 7, 4/5
Burnley IH (1995). Socioeconomic and spatial differentials in
mortality and means of committing suicide in New South
Wales, Australia. Social Science and Medicine 41, 687-698.
Busch KA, Fawcett J, Jacobs DG (2003). Clinical correlates of
inpatient suicide. Journal of Clinical Psychiatry 64, 14-
19.
Clarke RV, Lester D (1989). Suicide: Closing the exits. Springer
Verlag, New York.
Cooper J, Appleby L, Amons T (2002). Life events preceding
suicide by young people. Social Psychiatry and Psychiatric
Epidemiology 37, 271-275.
Cooper PN, Milroy CM (1994). Violent suicide in South
Yorkshire, England. Journal of Forensic Sciences 39, 657-
667.
De Leo D, Evans R, Neulinger K (2002). Hanging, firearm and
non-domestic gas suicides among males. Australian and
New Zealand Journal of Psychiatry 36, 183-189.
De Moore GM, Robertson AR (1999). Suicide attempts by
firearms and leaping from heights. American Journal of
Psychiatry 156, 1425-1431.
Hawton K, Fagg J, Simkin S, Harriss L, Malmberg A (1998).
Methods used for suicide by farmers in England and Wales.
British Journal of Psychiatry 173, 320-324.
Holden RR, Kerr PS, Mendonca JD, Velamoor VR (1998). Are
some motives more linked to suicide proneness than others?
Journal of Clinical Psychology 54, 569-576.
Holden RR, Kroner DG (2003). Differentiating suicide
motivations and manifestations in a forensic sample.
Canadian Journal of Behavioural Science 35, 35-44.
Johns D, Holden RR (1997). Differentiating suicidal motivations
and manifestations in a nonclinical population. Canadian
Journal of Behavioural Science 29, 266-274.
King EA, Baldwin DS, Sinclair JMA, Campbell MJ (2001). The
Wessex Recent In-Patient Study. British Journal of
Psychiatry 178, 537-542.
Lester D (1987). The perception of different methods of suicide.
Journal of General Psychology 115, 215-217.
Lester D (1993). Testosterone and suicide. Personality and
Individual Differences 15, 347-348.
Lester D (1997). Determinants of choice of method for suicide
and person/situation debate in psychology. Perceptual and
Motor Skills 85, 497-498.
Reith D, Whyte I, Carter G, McPherson M (2003). Adolescent
self-poisoning: A cohort study of subsequent suicide and
premature deaths. Crisis 24, 79-84.
Salib E, Maximous J (2002). Intimation of intent in elderly fatal
self-harm. International Journal of Psychiatry in Clinical
Practice 6, 155-161.
Scott KW (1994). Suicide in Wolverhampton. Medicine, Science
and the Law 34, 99-105.
Snowdon J, Baume P (2002). A study of suicides of older people
in Sydney. International Journal of Geriatric Psychiatry
17, 261-269.
Whelan G (2003). Alcohol: A much neglected risk factor in
elderly mental disorders. Current Opinion in Psychiatry
16, 609-614.