ELDERLY SUICIDE
AND PSYCHIATRIC MORBIDITY
Analysis of Coronors inquests of
200 cases of elderly suicide
in Cheshire 1989-2001
Honorary Senior Lecturer, Liverpool University
Consultant Psychiatrist, Hollins Park Hospital,
Warrington
Specialist Registrar in Old Age Psychiatry
North West Rotation
Introduction
It
appears that there is
an increasing interest among researchers on suicide amongst the elderly.
Many studies have attempted to explore the potential relationship
between the characteristics of elderly suicide victims and psychiatric
morbidity with inconsistent results in some respects. Elderly suicide
remains a problem of paramount importance because of the social and
psychological impact.
Objectives
In
this study we attempt to explore the characteristics of elderly suicide
victims who were known to have some form of psychiatric morbidity and
those who were not.
The
findings and possible clinical implications are discussed.
Methods
Data
was extracted from the records of coroner's inquests into all unexpected
deaths of persons aged 60 and over, in Cheshire over a period of 13 years
1989-2001. The Coroner's office covers the whole county of Cheshire
(1,000,000 of population). Verdicts of accidental death, misadventure and
open verdicts were not included in data collection.
Findings
The study showed no significant
difference between psychiatric morbidity and gender difference, living
alone, physical morbidity or marital status (P>0.05). Variables that
showed statistical significant difference between the two study groups;
included: childlessness, ethnic origin, previous contact with primary
services, evidence of intent, DSH and method of suicide.
Significantly more childless suicide
victims (n=71) suffered from a psychiatric illness (OR 0.6
95% 0.3 - 1.00
P<0.05). Elderly suicide victims who suffered with some form of
psychiatric morbidity, were significantly more likely to have contacted
their GP prior to their death
and to have had a history of deliberate
self harm (DSH ) but less likely to have shown any evidence of intent .
Elderly suicide victims who used a
violent method to terminate their lives, were significantly more likely to
be suffering from a psychiatric illness (OR 2
95% 1.2 - 3.7
P<0.05).
Conclusion
Elderly
suicide victims who did not suffer from any psychiatric illness are at
high risk of presenting with a first successful attempt with no prior
contact with their GP although they are likely to show an evidence of
intent. Early identification of this group and providing them with the
required support is a multi-agency task. Using effective diagnostic tools
in a primary care setting seems crucial.
Violent
suicidal methods are more likely to be used by mental illness sufferers.
Small
sample size and confounding are likely to have influenced the findings.
Further studies are needed to explore the characteristics of elderly
suicide victims and their suicidal acts in relation to utilisation of
relevant services.
BACKGROUND
It appears that there is an increasing
interest among researchers on suicide in
the elderly (De-Leo, 1997). With
the exception of Poland, the highest suicide rates are recorded in the
over-75 year olds in all countries which provided the World Health
Organisation with these statistics. Elderly
suicide remains a problem of paramount importance because of the social
and psychological impact it has on survivors.
[SURVIVORS? Just doesn't sound right. DW ]
Four research
approaches to studying putative risk factors in suicide among the elderly
were reviewed (Pearson, 1999). The
four risk factors were from epidemiological studies of suicidal
behaviours, clinic based follow-up studies, studies of suicide attempters
and psychological autopsy studies.
In addition
to social isolation, losses, physical illnesses and a past history of
suicide attempts, psychiatric illness was found to be one of the important
risk factors in suicide among the elderly.
This obviously has some implications for prevention of late life
suicide. Preliminary indications were that community outreach to
elders at risk and educational programmes for Primary Care providers on
the identification and treatment of late life depression, are effective at
lowering suicide rates (Cornwell, 1997).
The
epidemiology and psychiatric morbidity of the wish to be dead, suicidal
ideation and suicidal intent in a group of OLDER persons (aged over 70
years) were investigated (Barnow, 2001).
At the time of the study, 14.7% of the elderly community had
symptoms of tiredness of life, 5.4% wished to die and 1% showed suicidal
ideation or gestures.
Depending on
the intensity of suicide intent, 80% - 100% were clinically diagnosed as
suffering from psychiatric disorders and 50% - 75% showed symptoms which
fit the criteria of at least one specific psychiatric diagnosis.
Further, logistic regression analysis showed a significant
influence of major depression and specific diagnosis on suicidal intent in
old age.
Psychiatric
disorder and personality factors associated with suicide in older people
were investigated in a descriptive and case controlled study (Harwood,
2001). The study determined
the rates of psychiatric disorder and personality variables in a sample of
older people who had committed suicide and compared the rates in that
group of the sample with those in a control group of people who died from
natural causes in four countries and one large urban area in central
England. Depression, Personality Disorder and Personality Trait
Accentuation emerged as predictors of suicide in the case control
analysis.
One study
(Lynch, 2000) examined clinical and phenomenological correlates of
suicidal ideation among the elderly with Unipolar Depression.
The results
indicated that clinical variables associated with psychomotor retardation,
a history of Dysthymia, a previous psychiatric inpatient stay and being a
"younger" elder, were related to greater suicidal ideation which is
strongly associated with suicidal completion. The results also indicated
that feeling guilty, sinful or worthless, was associated with over six
times greater odds of having suicidal thoughts.
Another study
(Van-Exel, 2000) confirmed that elderly patients treated in clinical
psychiatry represent a group with more threatening and more disruptive
depressive illness.
This study
actually compared elderly patients with measured depression admitted to
the Psychiatric Hospital with those living in the community.
The two groups were compared with respect to demographic variables,
presenting symptoms, risk factors and treatment.
The characteristics that were found to be significantly more
prevalent in the clinical sample were late onset of depression, threat of
suicide, conflicts with significant others and use of antidepressant
medication. Chronic physical
illness was the only characteristic that was more prevalent in the
community sample.
Patients with
psychotic depression seem to have the least medical problems and those
with organic depression the most, whilst patients with minor depression
had the highest rate of family and marital problems, co-morbid personality
dysfunction and suicide attempts (Draper, 1997).
Patients with psychotic depression had the longest admission while
those with minor depression had the shortest.
The main diagnostic categories of depression used in this study
were: major depression, psychotic depression, minor depression and organic
depression.
Comprehensive
demographic and psychiatric data were available on 100 consecutive
referrals to a Liaison Psychiatric Service of elderly patients who
attempted suicide between 1989 and 1992 (Hepple, 1997).
This study also included detailed follow-up of survivors.
It appeared that elderly people who attempt suicide have a high
mortality, both from completed suicides and death from other causes.
The completed suicide rate was at least 1.5% per year, and the
repetition rate was 5.4% per year.
One study
examined the prevalence and correlates of dual diagnosis in older
psychiatric inpatient populations (Blixen, 1997).
Among the
main results of the study, was that significantly more dual diagnosis
elderly patients (17.7%) than mentally disordered patients (3.3%) made a
suicide attempt prior to admission to hospital.
As affective disorders in conjunction with alcohol abuse are the
most frequently found disorders in completed suicides, these findings seem
to support the routine use of diagnostic assessment and screening for both
substance abuse and mental disorders in elderly populations.
One review (Conwell,
1996) discussed how the profile of psychiatric diagnosis associated with
suicide varies across the life course.
It demonstrated that although affective disorders, substance abuse
disorder and their co-morbidity are common among suicide victims, the
relative prevalence of these factors differs as a function of age.
Whereas substance abuse is the most frequent diagnosis among
younger suicides, depressive illness is more closely associated with
suicide in the elderly.
Studies of
attempted suicide (parasuicide, deliberate self-harm) in the OLDER-age
population between 1985 and 1994 were critically reviewed with reference
to demography, suicide methods, stressors, psychiatric features and
outcome (Draper, 1996). The
studies seemed to consistently identify a number of factors long regarded
as being associated with suicidal behaviour in old age.
Mental illness was indeed one of those factors. Other factors
included being married [WIDOWED?], social isolation, impaired physical
health and high suicidal intent. Major
depression, obviously, was considered the main mental illness that is
associated with suicidal behaviour in old age.
In contrast to earlier research, recent studies have found
relationship problems to be a prominent factor.
This review has pointed out that the possibilities of the
psychological trait of hopelessness and the biological trait of low
central serotonergic activity being linked with suicide attempts in the
elderly requires further research.
RESULTS
This study included 200 suicide
recorded by the Cheshire Coroner between 1989 and 2001 of person aged 60
and above that reside within the County of Cheshire. The mean age of the
entire sample was 71 years (SD= 8, range 60-86). One hundred and seventeen
(58.5 % )of the sample were men and eighty three (41.5%) were women with
mean ages of 70 (SD 7) and 74 (SD 8) respectively.
Variables that showed statistical
significant difference between the two study groups (that are those with
and those without psychiatric morbidity); included: childlessness, ethnic
origin, previous contact with primary services, evidence of intent, DSH
and method of suicide.
The study showed no significant
difference between psychiatric morbidity and gender difference, living
alone, physical morbidity or marital status (P>0.05).
Significantly more childless suicide
victims (n=71) suffered from a psychiatric illness (OR 0.6
95% 0.3 - 1.00
P<0.05).
Not enough data WERE available on
psychiatric morbidity in relation to different ethnic origins.
Elderly suicide victims who suffered
with some form of psychiatric morbidity, were significantly more likely to
have contacted their GP at least once over a period of 3 months prior to
their death (OR 3.3 95% 1.8
- 5.9 P<0.00005).
Interestingly, suicide victims who
presented with an evidence of intent such as informing someone or leaving
a suicidal note were significantly less likely to have had a RECORDED
psychiatric illness (OR 0.6 95% 0.3 - 1.00 P<0.05).
History of deliberate self harm (DSH)
was found to be significantly higher (n=27) in suicide victims who had
psychiatric morbidity compared to those who were not known to be suffering
from a psychiatric illness (12) (OR 2.8
95% 1.3 - 5.9
P<0.05).
Understandably, those who were known to
psychiatric services were found to be more likely be suffering from some
form of psychiatric morbidity.
Elderly suicide victims who used a
violent method to terminate their lives, were significantly more likely to
be suffering from a psychiatric illness (OR 2
95% 1.2 - 3.7
P<0.05).
DISCUSSION
Interpretation of findings
About
half of our study sample was found to have suffered from some form of
psychiatric morbidity. Previous research indicated that 75% of the elderly
suicide have a psychiatric disorder at the time of death, most often
depression (63%). Personality disorder or personality trait accentuation
was present in 44% (Harwood, 2001). Whereas substance abuse is the most
frequent diagnosis among younger suicide, depressive illness was found by
some studies to be more closely associated with suicide in the elderly (Conwell,
1996), (Drapper, 1996).
It
was rather surprising, although interesting, to find no significant
correlation between psychiatric morbidity in one hand and gender
difference, living alone, physical morbidity, or marital status on the
other. One would have expected that the study will illustrate the possible
relationship between the occurrence of psychiatric illnesses and at least
some of those "potential risk factors".
It could, however, be argued that in
this particular sub-group, such factors can have a rather significant
effect on the course of the "already existent" mental illness, as
opposed to playing a direct aetiological role in its development. This
assumption is clearly not evidence-based.
"Do children protect against the
development of psychiatric illnesses amongst their parents?" is a
question that was highlighted by this study. Does the stress of not having
children play a role as a precipitating factor in the development of
mental illness? Do mental illnesses render their victims "unable" to
start a family?. One can only speculate.
The interesting paradox in this regard is the assumption that
children are likely to play a crucial role in attracting the attention of
relevant services to their parents' problems and hence getting them more
readily diagnosed of psychiatric illness than their childless
counterparts.
The lack of data on psychiatric
morbidity among victims of different ethnic origins does obviously point
out to a number of issues that are worth exploring by future studies. The
possible "genuine" epidemiological differences, the degree of service
utilisation among people with different cultural backgrounds and the
tendency to over or under diagnose certain problems in certain ethnic
subgroups; are only examples.
In spite of the usual assumption that
primary services play a major part in alerting secondary services to the
problems their clients might be suffering from, this seems to work both
ways. In other words, an elderly depressed patient who is known to
psychiatric services or primary services (by the mere fact that those were
the channels through which s/he was diagnosed in the first place) is more
likely to contact their GP at the time of crises. It is worth noting that
major depression, in the community, was thought to be more often
associated with chronic physical illness, which -in a routine GP
consultation can hamper the recognition and treatment of depression (Van-Exel).
This finding, however, does seem to be inconsistent among different
studies (Draper, 1997).
In some respect, helping our patients
to accept the potential usefulness of the medical model of depression and
other psychiatric illnesses can facilitate their contact with their GP
prior to making such a serious decision of putting an end to it all.
The finding that suicide victims who
showed some evidence of intent were less likely to have had a psychiatric
illness is interesting indeed.
This might denote that those who ended their lives through
committing a "rational suicide"-if this at all exists- have something
to "rationally" explain to their families as to why they chose such a
dramatic way of giving their "last say". Another argument would be
that a depressed elderly, by the very nature of his/her illness lack the
motivation and energy to give an explanation or reassurance to what they
perceive as the best or only way out, not only for them but also for their
care-givers.
Although this finding, in one hand,
points out to the fact that prevention of suicide is not, by any means, a
job that psychiatric profession should be exclusively hold accountable
for, missing out the diagnosis in both primary and secondary services
among this highly significant minority is a possibility that is always
worth baring in mind. Previous research has shown that suicidal ideation
in the elderly is usually a sign of a mental illness that warrants
diagnosis and treatment rather than assisted suicide (Barnow, 2001).
History of DSH is expected to be higher
among mental illness sufferers. Elderly patients who are at risk of
further self-harm were found, by other studies, to be more likely to be in
contact with Psychiatric services and to be suffering from persistent
depression (Hepple, 1997). Nevertheless, this finding can be of more
significance when we realise that those who have not had a formal
psychiatric diagnosis are likely to succeed first time.
Again, this illustrates the seriousness
of missing out this group and the importance of identifying them and
carrying out a through suicide risk assessment on them, baring in mind
that the lack of a clearly diagnosed psychiatric illness does not by any
means exclude the risk. It is also worth baring in mind that suicidal
ideation has been shown to be strongly associated suicide completion and
elders take their own lives more than any other age group (Lynch, 2000).
Although this is probably easier said
than done giving the absence of a major alarming factor; namely the mental
illness, it is yet crucial to take this seriously and accept the fact that
prevention of suicide in the elderly is a multi-agency task that is far
from being the job of a sole profession.
"Does violent suicide go hand in hand
with psychiatric morbidity" is a question that is highlighted by this
study. Although this could possibly be explained in some cases such as
psychotic illnesses, substance misuse or dementia, it is often hard to
explain the rationale behind carrying out a violent act by a retarded
flail elderly.
Limitations of the study
One
crucial limitation of this study is the lack of a matched control group
that consist of mentally ill elderly people who did not commit suicide.
The
other main limitations of this study include the small sample size and
confounding. The fact that the data were collected from only one county
might, at least partly, affect the external validity of some of the
findings should the sample be not representative to the elderly population
in the UK.
Collecting
information about the circumstances surrounding elderly suicide was rather
difficult giving their social isolation and/or reluctance to talk about
their emotional difficulties.
Although it
would have been useful to study different categories of psychiatric
disorders, this was not possible given the unavailability of the relevant
data beside the fact that such analysis was not the prime aim of this
study.
*Correspondence:
Hollins Park Hospital
Warrington
WA2 8WA
Tel: 01925 664123
Fax: 01925 664145
REFERENCES
Barnow
S, Linden M (2001) Epidemiology
and psychiatric morbidity of suicidal ideation among the elderly. Crisis
21 (4): 171-180
Blixen
C et al (1997) Dual diagnosis in elders discharged from a psychiatric
hospital. International-Journal-of-Geriatric-Psychiatry 12 (3): 307-313
Conwell
Y. (1997) Management of suicidal behaviour in the elderly. Psychiatric
Clinics of North America 20(3): 667-683
Conwell
Y, Brent D (1996) Suicide and aging I: Patterns of psychiatric diagnosis.
Suicide and aging: International perspectives. (pp. 15-30). New York, NY,
US: Springer Publishing Co, Inc. xvi, 235
De-Leo
D. (1997) Suicide in late at the end of the 1990s: A less neglected
topic? Crisis 18(2): 51-52
Draper
B, Anstey K (1997) Psychological stressors, physical illness and the
spectrum of depression in elderly inpatients.
Australian-and-New-Zealand-Journal-of-Psychiatry 30 (5): 567-572
Draper
B (1996) Attempted suicide in old age.
International-Journal-of-Geriatric-Psychiatry 11 (7): 577-587
Harwood
D et al (2001) Psychiatric disorder and personality factors associated with
suicide in older people: A descriptive and case-control study.
International-Journal-of-Geriatric-Psychiatry 16 (2): 155-165
Hepple
J, Quinton C (1997) One hundred cases of attempted suicide in the elderly
171: 42-46
Lynch
T et al (1999) Correlates of suicidal ideation among an elderly depressed
sample. Journal-of-Affective-Disorders 56 (1): 9-15
Pearson
J et al (1999) Studies
of suicide in later life: Methodologic considerations and research
directions. American-Journal-of-Geriatric-Psychiatry 7 (3): 203-210
Van-Exel
E et al(2000) The implication of selection bias in clinical studies of
late life depression: an impirical approach.
International-Journal-of-Geriatric-Psychiatry 15 (6) : 488-492
Table
Psychiatric Morbidity
No psychiatric Morbidity
Sex
Male
57
60
Female
42
41
OR 0.9 95% 0.5 - 1.6 P>0.05
Living
alone
Yes
50
56
No
49
45
OR 0.8 95% 0.5 - 1.4
P>0.05
Marital
status
Widowed
35
45
Others
64
56
OR 0.7 95% 0.4 - 1.4 P>0.05
Children
Yes
28
42
No
71
59
OR 0.6 95%
0.3 - 1.00 P<0.05
Ethnic
origin
White
99
95
Others
records not available
6
GP
contact
Yes
64
36
No
35
65
OR 3.3 95% 1.8 - 5.9
P<0.00005
Suicide
note
Yes
43
58
No
56
43
OR 0.6 95% 0.3 - 1.00 P<0.05
DSH
Yes
27
12
No
72
89
OR 2.8 95% 1.3 - 5.9
P<0.05
Physical
morbidity
Yes
63
64
No
36
37
OR 1.00 95%
0.6 - 1.8 P>0.05
Known
to services
Yes
55
2
No
44
99
Method
Violent
63
46
Non-violent
36
55
OR 2 95% 1.2 - 3.7
P<0.05
|
