Directory UMM :Data Elmu:jurnal:B:Biological Psichatry:Vol49.Issue2.2001:
Suicide and Alzheimer’s Pathology in the Elderly:
A Case–Control Study
Ana Rubio, Alice Lee Vestner, John M. Stewart, Nicholas T. Forbes,
Yeates Conwell, and Christopher Cox
Background: The single most important risk factor for
Alzheimer’s pathology is age. Elderly individuals are also
at increased risk for suicide, but comprehensive studies of
the association between Alzheimer’s pathology and suicide are lacking. We designed the current study to determine if Alzheimer’s disease changes are overrepresented
in elderly people committing suicide.
Methods: The design is a case– control study. Cases (n 5
28) were subjects older than 60 years of age who completed suicide. For each case, two age- and gendermatched individuals who died naturally were selected as
control subjects (n 5 56). Neuropathologic examination of
hippocampal sections was performed blindly and included
a modified Braak scoring system and semiquantitative
assessment of neurofibrillary tangles, amyloid deposition,
Lewy bodies, and Lewy-associated neurites. Data were
analyzed by conditional logistic regression.
Results: The brains of individuals who committed suicide
had higher modified Braak scores than those of matching
control subjects (p 5 .0028). The number of neurofibrillary tangles in CA1 was not an independent predictor of
suicide status in the statistical analysis (p 5 .16), although
the distribution was more highly skewed among the cases
(75th percentile of 10.5 for cases, vs. 2 for control
subjects).
Conclusions: Severe Alzheimer’s disease pathology is
overrepresented in elderly patients who complete suicide.
Biol Psychiatry 2001;49:137–145 © 2001 Society of Biological Psychiatry
Key Words: Suicide, elderly, Alzheimer’s disease, neurofibrillary pathology
From the Departments of Pathology and Laboratory Medicine (AR, ALV, JMS),
Psychiatry (YC), and Biostatistics (CC), University of Rochester School of
Medicine, and Office of the Monroe County Medical Examiner (NTF),
Rochester, New York.
Address reprint requests to Ana Rubio, M.D., Ph.D., University of Rochester
Medical Center, Dept. of Pathology (Neuropathology Unit), 601 Elmwood
Avenue, P.O. Box 626, Rochester NY 14642.
Received February 2, 2000; revised May 25, 2000; accepted May 31, 2000.
© 2001 Society of Biological Psychiatry
Introduction
T
he level of risk for suicide in patients with Alzheimer’s disease (AD) and other dementing illnesses,
although controversial, is generally considered low (Chiu
et al 1996; Draper et al 1998; Frierson 1991; Harris and
Barraclough 1997; Hepple and Quinton 1997; Lyness et al
1992). Depression (Wragg and Jeste 1989), aggressive
behavior (Aarsland et al 1996; Eastley and Wilcock 1997),
and occasional awareness of limitations, common in AD
(Kim and Hershey 1988), may be risk factors for suicide
(Blumenthal 1988; Margo and Finkel 1990; Willis 1987).
On the other hand, greater levels of supervision and
deficits in cognitive function may protect from suicide
(Conwell 1995). In our routine neuropathologic examination of brains from elderly individuals who committed
suicide, we found occasionally moderate to severe AD
pathology, but no data are available regarding the relative
incidence and severity of AD changes in the elderly
suicide population compared with gender- and agematched control subjects. A review of the literature
revealed only a few case reports of clinically demented
subjects who committed suicide and had neuropathologically confirmed AD changes (Lecso 1989; Rodhe et al
1995), but comprehensive studies of elderly suicides do
not commonly include neuropathologic evaluation (Conwell et al 1990; Ferris et al 1999; Margo and Finkel 1990).
Alzheimer’s-related neuropathologic changes are commonly found in the brains of older people. A recently
published multicenter study (Braak and Braak 1997) with
neuropathologic evaluation of 2661 subjects demonstrated
the presence of AD pathology in the majority of elderly
brains examined (more than 92% and 99% of subjects
older than 70 and 80 years, respectively, had some
neurofibrillary pathology). The severity and, most importantly, the topography of these changes follow a stereotyped pattern of progression with clearly defined stages
(Braak and Braak 1991; Fewster et al 1991; The National
Institute on Aging and Reagan Institute Working Group on
Diagnostic Criteria for the Neuropathological Assessment
of Alzheimer’s Disease 1997). Although the 1997 study
did not include clinical information on individuals as0006-3223/01/$20.00
PII S0006-3223(00)00952-5
138
A. Rubio et al
BIOL PSYCHIATRY
2001;49:137–145
sessed neuropathologically, other studies (Bancher et al
1996; Braak and Braak 1991; Braak et al 1993; Dickson
1997; Geddes et al 1996, 1997; Gertz et al 1996; Jellinger
and Bancher 1997; Masliah 1995; Newell et al 1999;
Xuereb et al 1995) have shown mixed correlation between
the pathologic and the clinical findings. The greatest
variation in cognitive status is among subjects with intermediate stages of AD pathology (limbic or Braak stages
III–IV), where the deficits range from not noticeable to
severe.
When the clinico-pathologic correlation is done by
grouping individuals according to their clinical status, the
results are equally heterogeneous. Of importance is the
wide range of AD neuropathology among individuals with
intact cognition (Davis et al 1999). Because the brunt of
the burden in the intermediate stages of AD pathology
localizes to the limbic system, alterations in affective
regulation and emotional behavior are expected to be
manifestations of these pathologic stages. It is thus plausible that mood changes or other emotional symptoms may
precede detectable cognitive decline in the clinical progression of the disease and may make subjects more
vulnerable to suicidal behavior.
We undertook this pilot project with the purpose of
determining if AD pathology was overrepresented in
elderly patients committing suicide.
available for 22 cases. Psychiatric diagnoses were determined by
administration of the Structured Clinical Interview for DSMIII-R (Spitzer et al 1986) to knowledgeable informants and
review of all available medical and psychiatric records. This
methodology is described in detail elsewhere (Conwell et al
1996). The reports were reviewed for Axis I diagnoses of
dementia, depression with and without psychosis, and substance
abuse.
Neuropathologic Examination
Cases were individuals over the age of 60 who completed
suicide. Subjects were identified from the files of the Monroe
County Medical Examiner’s Office. Other criteria for inclusion
were the performance of a complete autopsy and the availability
of hippocampal tissue sections.
Control subjects were chosen to match each case (two control
subjects per case) in age (within 3 years) and gender, and
included individuals who died naturally, had a routine autopsy
performed at the University of Rochester Medical Center, and
had hippocampal tissue available for neuropathologic examination. We selected the control subjects by reviewing the individual
front-sheets of the autopsy files to detect control subjects
matching each case for age and gender. Other factors, including
social, environmental, and medical, were not considered. The
presence of clinically diagnosed dementia and other chronic
illnesses, including neurologic disease, was not an exclusion
criterion for the selection of cases or control subjects.
Autopsy and clinical records were reviewed and the presence
of cancer as well as severe cardiac, respiratory, or other illnesses
was tabulated.
Seven-micron sections obtained at the caudal end of the anterior
hippocampus (generally at the level of the lateral geniculate
body) were obtained and stained with hematoxylin– eosin and the
modified Gallyas silver technique (Iqbal et al 1993). Tissue was
also immunostained with antibodies for phosphorylated tau
(paired helical filament [PHF] antibody, 1:20 dilution; kindly
donated by Dr. S.H. Yen), ubiquitin (Chemicon, Temecula, CA;
1:5000 dilution), and b-amyloid (Dako, Carpinteria, CA; 1:100
dilution following pretreatment of the tissue with formic acid).
Neuropathologic evaluation was performed by a single neuropathologist, blind to the case or control status and to all clinical
information. Neuropathologic staging was done by simplifying
the Braak and Braak grading system (Braak and Braak 1991) to
accommodate to hippocampus-only evaluation. A similar
method including the hippocampus and inferior temporal cortex
has been recently described and validated (Harding et al 2000).
In our study the majority of the sections included the hippocampus proper and the entorhinal and perirhinal cortices (beyond the
collateral sulcus). Individuals were characterized as stage 0 when
no neurofibrillary pathology was observed and stages I–II when
the neurofibrillary pathology was limited to the transentorhinal
region. Brains with mild entorhinal pathology and/or one to few
CA1 tangles at the junction of CA1 with the subiculum were also
included in this stage. Stages III–IV included cases with moderate involvement of CA1 and those with few neurofibrillary
tangles (NFTs) affecting CA4 of the hippocampus. In stages
V–VI the AD pathology was most extensive and intense,
affecting fascia dentata or CA2–3 of the hippocampus. The total
number of PHF-positive NFTs in CA1 of the hippocampus was
recorded for quantitative assessment of the intensity of Alzheimer’s-related pathology. Amyloid deposition in the cortex was
semiquantitatively evaluated (from 0, when there were no deposits, to 3, when the deposition was extensive) and the presence of
amyloid (congophilic) angiopathy recorded. The presence of
Lewy bodies and Lewy-associated neurites was determined by
the presence of ubiquitin-positive/tau-negative cytoplasmic inclusions in the deep parahippocampal gyrus and granular linear
deposits in the neuropil of the hippocampal CA3–2 area and
superficial layers of the parahippocampal cortex.
We limited the neuropathologic examination to hippocampal
sections because other brain areas were not uniformly available
for the cases.
Psychologic Autopsies
Statistical Analysis
As part of an ongoing study (“Suicide in later life: A psychological autopsy study”), psychological autopsy evaluations were
Since the study employed a matched design, the statistical
analysis was based on comparisons between cases and pairs of
Methods and Materials
Selection of Cases and Control Subjects
Suicide and Alzheimer’s in the Elderly
BIOL PSYCHIATRY
2001;49:137–145
Table 2. Suicide Method with Gender Distribution
Table 1. Demographic and Clinical Characteristics
Number
Age
Mean 6 SD
Range
Gender
Men
Women
Dementia
Other illnesses
Cancer
Cardiac
Respiratory
139
Cases
Control subjects
28
56
75.7 6 8.7
63–92
75.6 6 8.2
61–93
17 (60.7%)
11 (39.3%)
3 (10.7%)
34 (60.7%)
22 (39.3%)
7 (12.5%)
6 (21.4%)
7 (25%)
11 (39%)
18 (32.1%)
47 (83.9%)
21 (37.5%)
Method
Number (%)a
Drug overdoseb
Gunshot
CO intoxication
Strangulation (hanging)
Asphyxia (plastic bag)b
Drowning (bathtub)
Trauma (jumping)
8 (29%), 1:7
8 (29%), 7:1
6 (20%), 5:1
3 (10%), 2:1
2 (6%), 1:1
1 (3%), 0:1
1 (3%), 1:0
a
b
matched control subjects, using conditional logistic regression
models (McCullagh and Nelder 1989). The dependent variable in
each regression analysis was case/control status. Logistic regression models were used to examine associations with three
predictor variables: Braak stage (treated as a score with a range
of 0 –3 corresponding to Braak and Braak stages 0, I–II, III–IV,
and V–VI), number of NFTs in CA1, and amyloid score (0 –3).
The three predictor variables were examined separately and
combined in a single analysis.
Results
General Characteristics of the Subjects
Twenty-eight cases and 56 control subjects were included
in the study. Age, gender, presence of cognitive deficits,
and the incidence of pathologically confirmed chronic
cardiac, respiratory, or neoplastic illnesses are summarized in Table 1. Of note is the higher percentage of severe
cardiac pathology (atherosclerotic disease) among the
control subjects. The prevalence of cancer and severe
respiratory illnesses was comparable in the two groups.
Three (10.7%) cases and seven (12.5%) control subjects
had clinical dementia.
Demographics of Suicide
Twelve of the 28 cases (43%) wrote a note of intent. Three
other individuals (11%) had a conversation with a relative
or friend that retrospectively could be interpreted as
manifesting suicidal intent. The specific suicide method is
shown in Table 2. The two most common methods were
firearms, utilized by eight individuals (28.6%), seven of
whom were men, and drug overdose, employed by eight
subjects as well, seven of whom were women.
Twenty-two cases had a psychologic autopsy performed. Analysis of the Axis I evaluation is shown in
Table 3. Sixteen cases (73%) had a diagnosable affective
illness; five of the 16 had psychotic features. Eleven (50%)
were in their first episode of major unipolar depression,
Ratios are male:female.
One case combined drug overdose and plastic bag asphyxia.
three had a recurrent major depressive episode, and two
had dysthymia. Five subjects had an alcohol or drug use
disorder, and four had other diagnosable conditions
(schizophrenia [n 5 1], hypochondriasis [n 5 1], and
undifferentiated somatiform disorder [n 5 2]). A clinical
diagnosis of dementia was made in three subjects.
Brain Weight
Brain weight did not differ between cases and control
subjects, but the range was broader among the latter
(Figure 1A). Possible explanations for this broader range
are the higher incidence of brain edema among subjects
dying of chronic illnesses (responsible for the upper brain
weight value). Looking at the lower range, none of the
cases but three of the control subjects had brains weighing
less than 1000 g. The three were women with clinical
dementia. The neuropathologic substrate of the dementia
was AD in one case, Huntington’s disease (and mild AD
pathology) in another, and multiple brain infarcts (and
minimal AD pathology) in the third. As would be expected
and can be seen in Figure 1B, brain weight declined with
age (p 5 .035), with a gender difference (p , .0001).
Brain weight decreased as modified Braak stage score
increased (Figure 1C). The lack of apparent brain weight
decline among subjects within the most severe modified
Braak stage is most likely due to the increased representation of men (with their inherent heavier brains) in that
group (eight of 10 subjects were male).
Microscopic Studies
Figure 2 shows the distribution of cases and control
subjects among the four modified Braak score groups.
Although a similar proportion of both cases and control
subjects fell within the mild AD pathology (stages I–II),
29% of the cases, as compared with 4% of the control
subjects, were in the highest severity stages (V–VI).
Concomitantly, there were fewer cases than control subjects (4% vs. 20%) with no neurofibrillary pathology
(stage 0). The number of NFTs in CA1 was higher for
cases than for control subjects (Figure 3A), and the values
140
A. Rubio et al
BIOL PSYCHIATRY
2001;49:137–145
Table 3. Axis I Diagnosis (DSM-III-R) in 22 Cases with Psychological Autopsy
1
Dementia
Major depression, single
episode
Major depression, recurrent
Dysthymia
Accompanying psychosis
Alcohol dependence
Alcohol abuse
Drug abuse
Schizophrenia
Hypochondriasis
Undifferentiated somatiform
disorder
a
2
3
4
5
6
7
8
9
10
ea
f
f
f
f
f
f
f
11
f
Case
12
f
f
f
14
15
16
17
18
19
20
21
22
f
f
f
f
f
f
f
13
f
f
f
f
f
f
f
f
f
f
f
f
f
f
f
Dementia of Parkinson’s disease.
increase proportionally according to the modified Braak
stage (Figure 3B). The stage of amyloid deposition in the
cortex was comparable for the two groups (Table 4). Five
cases and 12 control subjects displayed amyloid angiopathy. One case and two control subjects had Lewy bodies in
the parahippocampal gyrus, and ubiquitin-positive granular neuropil staining in the superficial layers of the
parahippocampal gyrus was identified in 10 (36%) cases
and 20 (36%) control subjects.
Regression Analysis
The results of both the separate and combined logistic
regression analyses were completely consistent. A higher
Braak score predicted case status either alone (p 5 .0028)
or combined with the other two predictor variables (p 5
.014). This difference can be seen by comparing the
frequency distributions for the Braak scores in the two
groups (Figure 2). The cases clearly have relatively more
scores of 3 (Braak stages V–VI) and relatively fewer
scores of both 2 (Braak stages III–IV) and 0 (no neurofibrillary pathology). There was no evidence for an association between amyloid score and case/control status either
alone (p 5 .71) or combined with the other two variables
(p 5 .75). This is consistent with the similarity of the
distribution of scores in the two groups. The evidence for
an association with NFTs was marginal (p 5 .16). The
median numbers of NFTs were 2 for cases and 1 for
control subjects; however, the 75th percentile was 10.5 for
cases and 2 for control subjects, indicating a much more
skewed distribution among the cases. This kind of difference suggests that the assumption of a linear relationship
on the logit scale that is required by the logistic regression
model may be questionable for this variable.
Discussion
Our results show increased AD pathology in a population
of elderly people committing suicide compared with a
group of age- and gender-matched control subjects who
died of natural causes. The prevalence of AD neuropathology in the control group is comparable to that seen in other
population studies when stratified by age (Braak and
Braak 1997). Due to the nature of the dependent variable
being tested, suicide, the study has some intrinsic limitations, which we tried to minimize in our design. Specifically we addressed the issue of selection and ascertainment biases as follows. The selection of cases was based
on age, history of successful suicide, and availability of
brain tissue. The clinical or neurologic status did not play
a role in the performance of the autopsy, acquisition of
brain tissue, or selection for the study. Because all suicides
receive a forensic autopsy, we do not believe that social,
educational, or other factors could have biased the selection. Choosing appropriate control subjects is probably the
key factor for any case– control study. Our purpose was to
select a representative sample of the general population,
comparable in age and gender to that of the cases. To
increase statistical power, we chose two control subjects to
match each case. We believe that in our community
subjects receiving an autopsy at the university hospital are
more representative of the general population as a whole
than those having necropsies at the medical examiner’s
office (violent deaths, unexpected and unexplained deaths,
or deaths of individuals lacking physician attendance). It is
probably true that individuals dying at the university
hospital are sicker than those committing suicide (see also
Table 2), given that they all die of natural causes. We do
not believe that this factor biased the results. In fact, one
of the genetic risk factors for cardiovascular disease (the
Suicide and Alzheimer’s in the Elderly
BIOL PSYCHIATRY
2001;49:137–145
141
Figure 2. Distribution of cases (top) and control subjects (bottom) by modified Braak stage.
Figure 1. Brain weight by case/control status (A), age and
gender (B), and modified Braak stage (C). Data in A and C are
represented by box plots. The upper and lower limits of a box
represent the 75th and 25th percentiles, respectively, and the line
within the box indicates median values. Whiskers are representative of the range.
presence of one or more apolipoprotein E allele 4) is a risk
factor for AD as well. If our control subjects had a higher
prevalence of ApoE4, we could expect a higher degree of
AD pathology in the control group, which was not the
case. Because the neuropathologic grading was done
blindly to the case status and to all clinical information, a
biased assessment cannot explain the difference in AD
pathology between cases and control subjects. The neuropathologic examination was limited because of the lack of
extensive brain tissue available for examination among the
cases and the need to use the same methodology for the
control subjects. Although we would prefer to have more
brain areas available for testing, we believe that our
modified scoring system is representative of the degree of
AD pathology in the entire brain. This belief is substantiated by a recent study (Harding et al 2000) where the
authors evaluated 72 brains in different stages of AD
pathology and concluded that semiquantitative evaluation
of NFTs in the hippocampus and inferior temporal cortices
resulted in AD scores comparable to those obtained by the
full original Braak method (k statistic of .97).
Another factor limiting the scope of this study is the
lack of homogeneity of clinical data for cases and control
subjects. Although medical records were reviewed in both
groups, cases had additional information resulting from
the psychologic autopsy. As is common in case– control
studies like this one, where death is the key factor for
subject inclusion, the level of neuropsychologic evaluation
142
A. Rubio et al
BIOL PSYCHIATRY
2001;49:137–145
Figure 3. The number of neurofibrillary tangles (NFTs) in CA1
of the hippocampus plotted by case/control status (A) and by
modified Braak stage (B). Data are represented by box plots. The
upper and lower limits of a box represent the 75th and 25th
percentiles, respectively, and the line within the box indicates
median values. Whiskers are representative of the range.
was quite variable among individuals. Because of this, we
did not include clinical status as a dependent variable and
limited our end points to objective neuropathologic data
obtained in a blind fashion. Thus, our results do not
attempt to link suicide with AD (in the clinico-pathologic
sense), but with Alzheimer’s neuropathology.
The cases included in our study shared the common
experience of planning and carrying out an emotionally
and intellectually challenging act. Postmortem psychologic evaluation was done in 22 of the 28 cases included in
this study. Three individuals were identified as having
dementia. A 79-year-old man had dementia of Parkinson
disease. His neuropathologic evaluation revealed Dementia with Lewy bodies and mild AD changes. Secondly, a
Table 4. Amyloid Deposition in the Parahippocampal Cortex
No amyloid plaques
Mild amyloid deposits
Moderate amyloid deposits
Severe amyloid deposits
Cases [n (%)]
Control subjects
[n (%)]
16 (57)
6 (21)
5 (18)
1 (4)
26 (46)
14 (25)
8 (14)
8 (14)
75-year-old man with chronic schizophrenia was also
diagnosed as having dementia. His neuropathologic examination revealed mild AD pathology (Braak stages I–II),
hardly justifying this patient’s cognitive decline. Lastly,
the brain of a minimally demented 92-year-old female
nursing home resident had severe AD pathology (Braak
stages V–VI). Lack of clinical dementia in individuals
with severe AD pathology has been reported (Davis et al
1999). Alzheimer’s disease is a clinico-pathologic entity
(Berg and Morris 1994), and the currently recommended
algorithm for the pathologic confirmation of AD requires
the presence of clinical dementia (The National Institute
on Aging and Reagan Institute Working Group on Diagnostic Criteria for the Neuropathological Assessment of
Alzheimer’s Disease 1997). We interpret this apparent
lack of clinico-pathologic correlation as follows. Incipient
Alzheimer’s pathology is the rule rather than the exception
in the brains of elderly individuals. There is consensus
among experts that these pathologic changes are progressive, and that the progression advances in a stereotyped
manner according to brain topography. The tempo of the
progression is slow, with the disease taking years to
decades to advance from one Braak stage to another. The
tempo would vary among different subjects, depending on
known and unknown risk factors, both genetic (Arendt et
al 1997; Li et al 1997; Nitsch et al 1996) and environmental (Butler et al 1996; Itzhaki et al 1997). On the other
hand, individual susceptibility (brain reserve) plays a role
in the personal threshold for developing clinical dementia.
Because dementia is the key for the clinical diagnosis of
AD, we frequently observe neuropathologically comparable individuals with regard to Alzheimer’s pathology in
very different cognitive stages. Thus, although the majority of humans would probably develop in time enough
Alzheimer’s pathology to manifest clinical dementia, for
those in whom neuropathology starts late in life and
progresses slowly, or who have a larger “brain reserve,”
AD, in the clinico-pathologic sense, will not be manifest
within a normal life span. On the other hand, although
dementia is a cardinal feature of AD, it is not the only one,
and surely not the first to occur. Emotional-affective
symptoms most likely precede dementia, and because
cognitive decline is necessary for the clinical diagnosis of
AD, emotional symptoms due to Alzheimer’s pathology
would not be recognized as AD. In a recently published
study (Smith et al 2000) of individuals carrying the
Huntington disease mutation (so that the disease is genetically determined, provided patients live long enough), the
authors were able to detect subtle movement abnormalities
predicting disease progression. Similarly, in the limited
fashion allowed by the case– control method, we suggest
that suicide may be a manifestation of AD pathology.
Indeed, the population of patients at risk for suicide
Suicide and Alzheimer’s in the Elderly
because of such emotional changes may be larger than the
population who are ever diagnosed with AD.
In contrast to the low prevalence of clinical dementia
among the cases, postmortem psychological autopsy conferences demonstrated that the majority of individuals
studied (79%) had some type of depressive disorder (Table
3), most commonly single-episode major depression (50%
of the subjects with postmortem psychologic evaluation).
Neuropathologic examination for the presence of Alzheimer’s-related changes in those subjects with a single
episode of major depression revealed that nine of 11
individuals had early (Braak stages I–II) Alzheimer’s
pathology. Depressive symptoms in these patients with
early Alzheimer’s changes could either be a reaction to the
subjects’ awareness of their cognitive decline or a direct
manifestation of the entorhinal/limbic degeneration of
incipient (preclinical) AD. Alternatively, it could be argued that neurochemical imbalances usually seen in depressed patients committing suicide (for a review, see
Mann 1998) may stress neuronal systems, inducing neurodegeneration by decreasing plasticity and synaptic repair
(Brezun and Daszuta 1999). In fact, reduced activity of
5-hydroxytryptamine (5-HT) has been detected in AD
patients (Meltzer et al 1998; Terry et al 1991) and has been
linked to emotional imbalances, depression, and psychosis
(Zubenko et al 1991; Zweig et al 1988). These neurochemical imbalances may predispose patients to aggressive or
impulsive acts such as suicide, just as altered indices of
central serotonin function have been associated with suicide (Ågren 1980, 1983; Bachus et al 1997; Banki et al
1984; Brown et al 1982; Coccaro et al 1990; Mann and
Malone 1997; Mann et al 1996; Montgomery 1987; Ninan
et al 1984; Roy et al 1986; Stockmeier and Meltzer 1995;
Tra¨skman et al 1981; van Praag 1983; Virkkunen et al
1989) and other violent, aggressive behaviors (Virkkunen
et al 1987, 1995).
The majority of serotonin-containing neurons are located in the brain stem, predominantly in the dorsal and
ventral raphe (Frazer and Hensler 1999). The low serotonergic status seen in suicide victims does not seem to be
due to fewer neurons in the raphe nuclei (Underwood et al
1999). On the other hand, medial raphe neurons, including
serotonergic neurons, have NFTs in AD (Halliday et al
1992). Axons from the median raphe neurons project to
the hippocampus and cerebral cortex, whereas from the
dorsal raphe they reach the striatum and cortex. Numerous
pre- and postsynaptic serotonin receptors have been identified and their genes mapped. Polymorphisms for 5-HT2A
and 5-HT2C may be responsible not only for susceptibility
to psychosis (Gutierrez et al 1996; Williams et al 1997),
but also for the psychotic symptomatology in patients with
AD (Holmes et al 1998). These findings may be gender
dependent, as indicated in animal models by the influence
BIOL PSYCHIATRY
2001;49:137–145
143
of aromatase conversion of androgens to estrogens in the
expression of 5-HT2A and serotonin transporters in the
brain (Fink et al 1999).
A third possible explanation of the association that we
have observed is that both suicidal behavior and Alzheimer’s pathology may be independently linked to a genetic
or environmental factor that affects both (Li et al 1997;
Oliveira et al 1998).
Our conclusion that people with moderate to severe AD
pathology are overrepresented in a sample of elderly
patients who completed suicide should be validated in a
larger, prospective study with psychologic autopsy data
for control subjects as well as for cases. The psychologic
examination should include cognitive, emotional, and
behavioral assessment. We hypothesize that the emotional-affective state of cognitively intact or minimally impaired individuals who later show AD symptoms may put
them at risk for suicide before they reach clinically
diagnosable AD.
Supported in part by grants AG-08665 and RO1 MH 54682.
The authors thank Ms. Frances Vito and Mr. Matt Frank for the
histochemical and immunohistochemical preparations.
Presented at the “Unique Features of Neuropsychiatric Disease in the
Elderly” panel, annual meeting of the American College of Neuropsychopharmacology, December 12–16, 1999, Acapulco, Mexico.
References
Aarsland D, Cummings JL, Yenner G, Miller B (1996): Relationship of aggressive behavior to other neuropsychiatric
symptoms in patients with Alzheimer’s disease. Am J Psychiatry 153:243–247.
Ågren H (1980): Symptom patterns in unipolar and bipolar
depression correlating with monoamine metabolites in the
cerebrospinal fluid: I. General patterns. Psychiatry Res
3:211–223.
Ågren H (1983): Life at risk: Markers of suicidality in depression. Psychiatry Dev 1:87–103.
Arendt T, Schindler C, Bruckner MK, Eschrich K, Bigl V,
Zedlick D, Marcova L (1997): Plastic neuronal remodeling is
impaired in patients with Alzheimer’s disease carrying apolipoprotein epsilon 4 allele. J Neurosci 17:516 –529.
Bachus SE, Hyde TM, Akil M, Shannon Weikert C, Vawter MP,
Kleinman JE (1997): Neuropathology of suicide: A review
and an approach. Ann N Y Acad Sci 836:201–219.
Bancher C, Jellinger K, Lassmann H, Fischer P, Leblhuber F
(1996): Correlations between mental state and quantitative
neuropathology in the Vienna Longitudinal Study of Dementia. Eur J Psychiatry Clin Neurosci 246:137–146.
Banki CM, Arato M, Papp Z, Kurcz M (1984): Biochemical
markers in suicidal patients. Investigations with cerebrospinal
fluid amine metabolites and neuroendocrine tests. J Affect
Disord 6:341–350.
Berg L, Morris JC (1994): Diagnosis. In: Terry RD, Katzman R,
144
BIOL PSYCHIATRY
2001;49:137–145
Bick KL, editors. Alzheimer’s Disease. New York: Raven,
9 –25.
Blumenthal SJ (1988): Suicide: A guide to risk factors, assessment and treatment of suicidal patients. Med Clin North Am
72:937–971.
Braak H, Braak E (1991): Neuropathological staging of Alzheimer-related changes. Acta Neuropathol 82:239 –259.
Braak H, Braak E (1997): Frequency of stages of Alzheimerrelated lesions in different age categories. Neurobiol Aging
18:351–357.
Braak H, Duyckaerts C, Braak E, Piette F (1993): Neuropathological staging of Alzheimer-related changes correlates with
psychometrically assessed intellectual status. In: Corain B,
Iqbal K, Nicolini M, Wiblad B, Wisniewski H, Zatta P,
editors. Alzheimer’s Disease: Advances in Clinical and Basic
Research. Chichester, UK: Wiley, 131–137.
Brezun JM, Daszuta A (1999): Depletion in serotonin decreases
neurogenesis in the dentate gyrus and the subventricular zone
of adult rats. Neuroscience 89:999 –1002.
Brown GL, Ebert MH, Goyer PF, Jimersom DC, Klein WJ,
Bunney WE, et al (1982): Aggression, suicide and serotonin:
Relationship to CSF amine metabolites. Am J Psychiatry
139:741–746.
Butler SM, Ashford JW, Snowdon DA (1996): Age, education,
and changes in the Mini-Mental State Exam scores of older
women: Findings from the Nun Study. J Am Geriatr Soc
44:675– 681.
Chiu HF, Lam LC, Pang AH, Leung CM, Wong CK (1996):
Attempted suicide by Chinese elderly in Hong Kong. Gen
Hosp Psychiatry 18:444 – 447.
Coccaro EF, Silver LJ, Owen K, Davis KL (1990): Serotonin in
mood and personality disorders. In: Coccaro EF, Murphy DL,
editors. Serotonin in Major Psychiatric Disorders. Washington, DC: APA Press, 71–97.
Conwell Y (1995): Dementia. Crisis 16:5– 6.
Conwell Y, Duberstein PR, Cox C, Herrmann JH, Forbes NT,
Caine ED (1996): Relationships of age and axis I diagnoses in
victims of completed suicide: A psychological autopsy study.
Am J Psychiatry 153:1001–1008.
Conwell Y, Rotenberg M, Caine ED (1990): Completed suicide
at age 50 and over. J Am Geriatr Soc 38:640 – 644.
Davis DG, Schnitt FA, Wekstein DR, Markesbery WR (1999):
Alzheimer neuropathologic alterations in aged cognitively
normal subjects. J Neuropathol Exp Neurol 58:376 –388.
Dickson DW (1997): The value of cross sectional neuroanatomical studies as a conceptual framework for prospective clinicopathological studies. Neurobiol Aging 18:382–386.
Draper B, Moore CM, Brodaty H (1998): Suicidal ideation and
the “wish to die” in dementia patients: The role of depression.
Age Ageing 27:503–507.
Eastley R, Wilcock GK (1997): Prevalence and correlates of
aggressive behaviours occurring in patients with Alzheimer’s
disease. Int J Geriatr Psychiatry 12:484 – 487.
Ferris SH, Hofeldt GT, Carbone G, Masciandaro P, Troetel WM,
Imbimbo BP (1999): Suicide in two patients with a diagnosis
of probable Alzheimer disease. Alzheimer Dis Assoc Disord
13:88 –90.
Fewster PH, Griffin-Brooks S, MacGregor J, Ojalvo-Rose E,
A. Rubio et al
Ball MJ (1991): A topographical pathway by which histopathological lesions disseminate through the brain of patients with Alzheimer’s disease. Dementia 2:121–132.
Fink G, Sumner B, Rosie R, Wilson H, McQueen J (1999):
Androgen actions on central serotonin neurotransmission:
relevance for mood, mental state and memory. Behav Brain
Res 105:53– 68.
Frazer A, Hensler JG (1999): Serotonin. In: Siegel GJ, editor.
Basic Neurochemistry. Molecular, Cellular and Medical
Aspects, 6th ed. Philadelphia: Lippincott, Williams and
Wilkins, 263–292.
Frierson RL (1991): Suicide attempts by the old and the very old.
Arch Intern Med 151:141–144.
Geddes JW, Snowdon DA, Soultanian NS, Tekirian TL, Riley
KP, Ashford JW, et al (1996): Braak stages III-IV of
Alzheimer-related neuropathology are associated with mild
memory loss, stages V-VI are associated with dementia:
Findings from the Nun study. J Neuropathol Exp Neurol
55:617.
Geddes JW, Tekirian TL, Soultanian NS, Ashford JW, Davis
DG, Markesbery WR (1997): Comparison of neuropathologic
criteria for the diagnosis of Alzheimer’s disease. Neurobiol
Aging 18(suppl 4):S99 –S105.
Gertz HJ, Xuereb JH, Huppert FA, Brayne C, Kruger H, McGee
MA, et al (1996): The relationship between clinical dementia
and neuropathological staging (Braak) in a very elderly
community sample. Eur Arch Psychiatry Clin Neurosci 246:
132–136.
Gutierrez B, Fananas L, Arranz MJ, Valles V, Guillamat R, van
Os J, Collier D (1996): Allelic association analysis of the
5-HT2C receptor gene in bipolar disorder. Neurosci Lett
212:65– 67.
Halliday GM, McCann HL, Pamphlett R, Brooks WS, Creasey
H, McCusker E, et al (1992): Brain stem serotonin-synthesizing neurons in Alzheimer’s disease: A clinicopathological
correlation. Acta Neuropathol 84:638 – 650.
Harding AJ, Kril JJ, Halliday GM (2000): Practical measures to
simplify the Braak tangle staging method for routine pathological screening. Acta Neuropathol 99:199 –208.
Harris EC, Barraclough B (1997): Suicide as an outcome for
mental disorders: A meta-analysis. Br J Psychiatry 170:205–
228.
Hepple J, Quinton C (1997): One hundred cases of attempted
suicide in the elderly. Br J Psychiatry 171:42– 46.
Holmes C, Arranz MJ, Powell JF, Collier DA, Lovestone S
(1998): 5-HT2A and 5-HT2C receptor polymorphisms and
psychopathology in late onset Alzheimer’s disease. Hum Mol
Genet 7:1507–1509.
Iqbal K, Braak H, Braak E, Grundke-Iqbal I (1993): Silver
labeling of Alzheimer neurofibrillary changes and brain b
amyloid. J Histotechnol 16:335–342.
Itzhaki RF, Lin WR, Shang D, Wilcock GK, Faragher B,
Jamieson GA (1997): Herpes simplex virus type 1 in brain
and risk of Alzheimer’s disease. Lancet 349:241–244.
Jellinger KA, Bancher C (1997): Proposals for re-evaluation of
current autopsy criteria for the diagnosis of Alzheimer’s
disease. Neurobiol Aging 18(suppl 4):S55–S65.
Kim KY, Hershey LA (1988): Diagnosis and treatment of
depression in the elderly. Int J Psychiatry Med 18:211–221.
Suicide and Alzheimer’s in the Elderly
Lecso PA (1989): Murder-suicide in Alzheimer’s disease. J Am
Geriatr Soc 37:167–168.
Li T, Homes C, Sham PC, Vallada H, Birkett J, Kirov G, et al
(1997): Allelic functional variation of serotonin transporter
expression is a susceoptibility factor for late onset Alzheimer’s disease. Neuroreport 8:683– 686.
Lyness JM, Conwell Y, Nelson JC (1992): Suicide attempts in
elderly psychiatric inpatients. J Am Geriatr Soc 40:320 –324.
Mann JJ (1998): The neurobiology of suicide. Nat Med 1998;4:
25–30.
Mann JJ, Malone KM (1997): Cerebrospinal fluid amines and
higher lethality suicide attempts in depressed inpatients. Biol
Psychiatry 41:162–171.
Mann JJ, Underwood MD, Arango V (1996): Postmortem studies
of suicide victims. In: Watson SJ, editor. Biology of Schizophrenia and Affective Disease, 1st ed. Washington, DC: APP,
197–220.
Margo GM, Finkel JA (1990): Early dementia as a risk factor for
suicide. Hosp Community Psychiatry 41:676 – 678.
Masliah E (1995): Natural evolution of the neurodegenerative
alterations in Alzheimer’s disease. Neurobiol Aging 16:280 –
282.
McCullagh P, Nelder JA (1989): Generalized Linear Models,
2nd ed. London: Chapman and Hall.
Meltzer CC, Smith G, DeKosky ST, Pollock BG, Mathis CA,
Moore RY, et al (1998): Serotonin in aging, late-life depression, and Alzheimer’s disease: The emerging role of functional imaging. Neuropsychopharmacology 18:407– 430.
Montgomery SA (1987): The psychopharmacology of borderline
personality disorders. Acta Psychiatr Belg 87:260 –266.
The National Institute on Aging, Reagan Institute Working
Group on Diagnostic Criteria for the Neuropathological
Assessment of Alzheimer’s Disease (1997): Consensus recommendation for the postmortem diagnosis of Alzheimer’s
disease. Neurobiol Aging 18(suppl 1):S1–S2.
Newell KL, Hyman BT, Growdon JH, Hedley-Whyte ET (1999):
Application of the National Institute on Aging (NIA)-Reagan
Institute criteria for the neuropathologic diagnosis of Alzheimer disease. J Neuropathol Exp Neurol 58:1147–1155.
Ninan PT, van Kammen DP, Scheinin M, Linnoila M, Bunney
WE, Goodwin FK (1984): CSF 5-hydroxyindoleacetic acid
levels in suicidal schizophrenic patients. Am J Psychiatry
141:566 –569.
Nitsch RM, Deng M, Growdon JH, Wurtman RJ (1996): Serotonin 5-HT2a and 5-HT2c receptors stimulate amyloid precursor protein ectodomain secretion. J Biol Chem 271:4188 –
4194.
Oliveira JR, Gallindo RM, Maia LG, Brito-Marques PR, Otto
PA, Passos-Bueno MR, et al (1998): The short variant of the
polymorphism within the promoter region of the serotonin
transporter gene is a risk factor for late onset Alzheimer’s
disease. Mol Psychiatry 3:438 – 451.
Rodhe K, Peskind ER, Raskind MA (1995): Suicide in two
patients with Alzheimer’s disease. J Am Geriatr Soc 43:187–
189.
Roy A, Ågren H, Pickar D, Linnoila M, Doran A, Cutler N, et al
BIOL PSYCHIATRY
2001;49:137–145
145
(1986): Reduced CSF concentrations of homovanilic acid and
homovanilic acid to 5-hydroxyindoleacetic acid ratios in
depressed patients: relationship to suicidal behavior and
dexamethasone nonsuppression. Am J Psychiatry 143:1539 –
1545.
Smith MA, Brandt J, Shadmehr R (2000): Motor disorder in
Huntington’s disease begins as a dysfunction in error feedback control. Nature 403:544 –549.
Spitzer RL, Williams JBW, Gibbon M (1986): Structured Clinical Interview for DSM-IIIR (SCID). New York: New York
State Psychiatric Institute, Biometrics Research.
Stockmeier CA, Meltzer HY (1995): Biologic studies of suicide
in schizophrenia. Clin Neuropharmacol 18(3):S9 –S17.
Terry RD, Masliah E, Salmon DP, Butters N, DeTeresa R, Hill
R, et al (1991): Physical basis of cognitive alterations in
Alzheimer’s disease: Synapse loss is the major correlate of
cognitive impairment. Ann Neurol 30:572–580.
Tra¨skman L, Åsberg M, Bertilsson L, Sjo¨strand L (1981):
Monoamine metabolites in CSF and suicidal behavior. Arch
Gen Psychiatry 38:631– 636.
Underwood MD, Khaibulina AA, Ellis SP, Moran A, Rice PM,
Mann JJ, Arango V (1999): Morphometry of the dorsal raphe
nucleus serotonergic neurons in suicide victims. Biol Psychiatry 46:473– 483.
van Praag HM (1983): CSF 5-HIAA and suicide in nondepressed schizophrenics. Lancet ii:977–978.
Virkkunen M, De Jong J, Bartko J, Linnoila M (1989): Psychobiological concomitants of history of suicide attempts among
violent offenders and impulsive fire setters. Arch Gen Psychiatry 46:604 – 606.
Virkkunen M, Goldman D, Nielsen D, Linnoila M (1995): Low
brain serotonin turnover rate (low CSF 5-HIAA) and impulsive violence. J Psychiatr Neurosci 20:271–275.
Virkkunen M, Nuutila A, Goodwin FK, Linnoila M (1987):
Cerebrospinal fluid monoamine metabolite levels in male
arsonists. Arch Gen Psychiatry 44:241–247.
Williams J, McGuffin P, Nothen MM, Owen MJ, EMASS
Consortium (1997): Meta-analysis of association between the
5-HT2a receptor T102C polymorphism and schizophrenia.
Lancet 349:1221.
Willis RH (1987): Suicide risk in elderly persons: Diagnosis and
management. Mt Sinai J Med 54:14 –17.
Wragg RE, Jeste DV (1989): Overview of depression and
psychosis in Alzheimer’s disease. Am J Psychiatry 146:577–
587.
Xuereb JH, Gertz HJ, Huppert F, Brayne C, Wischik CM,
Mukaetoba-Ladinska E (1995): The application of Braak’s
staging model of Alzheimer-type pathology to neuropathological diagnosis of dementia. Neuropathol Appl Neurobiol
21:44.
Zubenko GS, Moosy J, Martinez AJ, Rao G, Claassen D, Rosen
J, Kopp U (1991): Neuropathologic and neurochemical correlates of psychosis in primary dementia. Arch Neurol 48:
619 – 624.
Zweig RM, Ross CA, Hedreen JC, Steele C, Cardillo JE,
Whitehouse PJ, et al (1988): The neuropathology of aminergic nuclei in Alzheimer’s disease. Ann Neurol 24:233–242.
A Case–Control Study
Ana Rubio, Alice Lee Vestner, John M. Stewart, Nicholas T. Forbes,
Yeates Conwell, and Christopher Cox
Background: The single most important risk factor for
Alzheimer’s pathology is age. Elderly individuals are also
at increased risk for suicide, but comprehensive studies of
the association between Alzheimer’s pathology and suicide are lacking. We designed the current study to determine if Alzheimer’s disease changes are overrepresented
in elderly people committing suicide.
Methods: The design is a case– control study. Cases (n 5
28) were subjects older than 60 years of age who completed suicide. For each case, two age- and gendermatched individuals who died naturally were selected as
control subjects (n 5 56). Neuropathologic examination of
hippocampal sections was performed blindly and included
a modified Braak scoring system and semiquantitative
assessment of neurofibrillary tangles, amyloid deposition,
Lewy bodies, and Lewy-associated neurites. Data were
analyzed by conditional logistic regression.
Results: The brains of individuals who committed suicide
had higher modified Braak scores than those of matching
control subjects (p 5 .0028). The number of neurofibrillary tangles in CA1 was not an independent predictor of
suicide status in the statistical analysis (p 5 .16), although
the distribution was more highly skewed among the cases
(75th percentile of 10.5 for cases, vs. 2 for control
subjects).
Conclusions: Severe Alzheimer’s disease pathology is
overrepresented in elderly patients who complete suicide.
Biol Psychiatry 2001;49:137–145 © 2001 Society of Biological Psychiatry
Key Words: Suicide, elderly, Alzheimer’s disease, neurofibrillary pathology
From the Departments of Pathology and Laboratory Medicine (AR, ALV, JMS),
Psychiatry (YC), and Biostatistics (CC), University of Rochester School of
Medicine, and Office of the Monroe County Medical Examiner (NTF),
Rochester, New York.
Address reprint requests to Ana Rubio, M.D., Ph.D., University of Rochester
Medical Center, Dept. of Pathology (Neuropathology Unit), 601 Elmwood
Avenue, P.O. Box 626, Rochester NY 14642.
Received February 2, 2000; revised May 25, 2000; accepted May 31, 2000.
© 2001 Society of Biological Psychiatry
Introduction
T
he level of risk for suicide in patients with Alzheimer’s disease (AD) and other dementing illnesses,
although controversial, is generally considered low (Chiu
et al 1996; Draper et al 1998; Frierson 1991; Harris and
Barraclough 1997; Hepple and Quinton 1997; Lyness et al
1992). Depression (Wragg and Jeste 1989), aggressive
behavior (Aarsland et al 1996; Eastley and Wilcock 1997),
and occasional awareness of limitations, common in AD
(Kim and Hershey 1988), may be risk factors for suicide
(Blumenthal 1988; Margo and Finkel 1990; Willis 1987).
On the other hand, greater levels of supervision and
deficits in cognitive function may protect from suicide
(Conwell 1995). In our routine neuropathologic examination of brains from elderly individuals who committed
suicide, we found occasionally moderate to severe AD
pathology, but no data are available regarding the relative
incidence and severity of AD changes in the elderly
suicide population compared with gender- and agematched control subjects. A review of the literature
revealed only a few case reports of clinically demented
subjects who committed suicide and had neuropathologically confirmed AD changes (Lecso 1989; Rodhe et al
1995), but comprehensive studies of elderly suicides do
not commonly include neuropathologic evaluation (Conwell et al 1990; Ferris et al 1999; Margo and Finkel 1990).
Alzheimer’s-related neuropathologic changes are commonly found in the brains of older people. A recently
published multicenter study (Braak and Braak 1997) with
neuropathologic evaluation of 2661 subjects demonstrated
the presence of AD pathology in the majority of elderly
brains examined (more than 92% and 99% of subjects
older than 70 and 80 years, respectively, had some
neurofibrillary pathology). The severity and, most importantly, the topography of these changes follow a stereotyped pattern of progression with clearly defined stages
(Braak and Braak 1991; Fewster et al 1991; The National
Institute on Aging and Reagan Institute Working Group on
Diagnostic Criteria for the Neuropathological Assessment
of Alzheimer’s Disease 1997). Although the 1997 study
did not include clinical information on individuals as0006-3223/01/$20.00
PII S0006-3223(00)00952-5
138
A. Rubio et al
BIOL PSYCHIATRY
2001;49:137–145
sessed neuropathologically, other studies (Bancher et al
1996; Braak and Braak 1991; Braak et al 1993; Dickson
1997; Geddes et al 1996, 1997; Gertz et al 1996; Jellinger
and Bancher 1997; Masliah 1995; Newell et al 1999;
Xuereb et al 1995) have shown mixed correlation between
the pathologic and the clinical findings. The greatest
variation in cognitive status is among subjects with intermediate stages of AD pathology (limbic or Braak stages
III–IV), where the deficits range from not noticeable to
severe.
When the clinico-pathologic correlation is done by
grouping individuals according to their clinical status, the
results are equally heterogeneous. Of importance is the
wide range of AD neuropathology among individuals with
intact cognition (Davis et al 1999). Because the brunt of
the burden in the intermediate stages of AD pathology
localizes to the limbic system, alterations in affective
regulation and emotional behavior are expected to be
manifestations of these pathologic stages. It is thus plausible that mood changes or other emotional symptoms may
precede detectable cognitive decline in the clinical progression of the disease and may make subjects more
vulnerable to suicidal behavior.
We undertook this pilot project with the purpose of
determining if AD pathology was overrepresented in
elderly patients committing suicide.
available for 22 cases. Psychiatric diagnoses were determined by
administration of the Structured Clinical Interview for DSMIII-R (Spitzer et al 1986) to knowledgeable informants and
review of all available medical and psychiatric records. This
methodology is described in detail elsewhere (Conwell et al
1996). The reports were reviewed for Axis I diagnoses of
dementia, depression with and without psychosis, and substance
abuse.
Neuropathologic Examination
Cases were individuals over the age of 60 who completed
suicide. Subjects were identified from the files of the Monroe
County Medical Examiner’s Office. Other criteria for inclusion
were the performance of a complete autopsy and the availability
of hippocampal tissue sections.
Control subjects were chosen to match each case (two control
subjects per case) in age (within 3 years) and gender, and
included individuals who died naturally, had a routine autopsy
performed at the University of Rochester Medical Center, and
had hippocampal tissue available for neuropathologic examination. We selected the control subjects by reviewing the individual
front-sheets of the autopsy files to detect control subjects
matching each case for age and gender. Other factors, including
social, environmental, and medical, were not considered. The
presence of clinically diagnosed dementia and other chronic
illnesses, including neurologic disease, was not an exclusion
criterion for the selection of cases or control subjects.
Autopsy and clinical records were reviewed and the presence
of cancer as well as severe cardiac, respiratory, or other illnesses
was tabulated.
Seven-micron sections obtained at the caudal end of the anterior
hippocampus (generally at the level of the lateral geniculate
body) were obtained and stained with hematoxylin– eosin and the
modified Gallyas silver technique (Iqbal et al 1993). Tissue was
also immunostained with antibodies for phosphorylated tau
(paired helical filament [PHF] antibody, 1:20 dilution; kindly
donated by Dr. S.H. Yen), ubiquitin (Chemicon, Temecula, CA;
1:5000 dilution), and b-amyloid (Dako, Carpinteria, CA; 1:100
dilution following pretreatment of the tissue with formic acid).
Neuropathologic evaluation was performed by a single neuropathologist, blind to the case or control status and to all clinical
information. Neuropathologic staging was done by simplifying
the Braak and Braak grading system (Braak and Braak 1991) to
accommodate to hippocampus-only evaluation. A similar
method including the hippocampus and inferior temporal cortex
has been recently described and validated (Harding et al 2000).
In our study the majority of the sections included the hippocampus proper and the entorhinal and perirhinal cortices (beyond the
collateral sulcus). Individuals were characterized as stage 0 when
no neurofibrillary pathology was observed and stages I–II when
the neurofibrillary pathology was limited to the transentorhinal
region. Brains with mild entorhinal pathology and/or one to few
CA1 tangles at the junction of CA1 with the subiculum were also
included in this stage. Stages III–IV included cases with moderate involvement of CA1 and those with few neurofibrillary
tangles (NFTs) affecting CA4 of the hippocampus. In stages
V–VI the AD pathology was most extensive and intense,
affecting fascia dentata or CA2–3 of the hippocampus. The total
number of PHF-positive NFTs in CA1 of the hippocampus was
recorded for quantitative assessment of the intensity of Alzheimer’s-related pathology. Amyloid deposition in the cortex was
semiquantitatively evaluated (from 0, when there were no deposits, to 3, when the deposition was extensive) and the presence of
amyloid (congophilic) angiopathy recorded. The presence of
Lewy bodies and Lewy-associated neurites was determined by
the presence of ubiquitin-positive/tau-negative cytoplasmic inclusions in the deep parahippocampal gyrus and granular linear
deposits in the neuropil of the hippocampal CA3–2 area and
superficial layers of the parahippocampal cortex.
We limited the neuropathologic examination to hippocampal
sections because other brain areas were not uniformly available
for the cases.
Psychologic Autopsies
Statistical Analysis
As part of an ongoing study (“Suicide in later life: A psychological autopsy study”), psychological autopsy evaluations were
Since the study employed a matched design, the statistical
analysis was based on comparisons between cases and pairs of
Methods and Materials
Selection of Cases and Control Subjects
Suicide and Alzheimer’s in the Elderly
BIOL PSYCHIATRY
2001;49:137–145
Table 2. Suicide Method with Gender Distribution
Table 1. Demographic and Clinical Characteristics
Number
Age
Mean 6 SD
Range
Gender
Men
Women
Dementia
Other illnesses
Cancer
Cardiac
Respiratory
139
Cases
Control subjects
28
56
75.7 6 8.7
63–92
75.6 6 8.2
61–93
17 (60.7%)
11 (39.3%)
3 (10.7%)
34 (60.7%)
22 (39.3%)
7 (12.5%)
6 (21.4%)
7 (25%)
11 (39%)
18 (32.1%)
47 (83.9%)
21 (37.5%)
Method
Number (%)a
Drug overdoseb
Gunshot
CO intoxication
Strangulation (hanging)
Asphyxia (plastic bag)b
Drowning (bathtub)
Trauma (jumping)
8 (29%), 1:7
8 (29%), 7:1
6 (20%), 5:1
3 (10%), 2:1
2 (6%), 1:1
1 (3%), 0:1
1 (3%), 1:0
a
b
matched control subjects, using conditional logistic regression
models (McCullagh and Nelder 1989). The dependent variable in
each regression analysis was case/control status. Logistic regression models were used to examine associations with three
predictor variables: Braak stage (treated as a score with a range
of 0 –3 corresponding to Braak and Braak stages 0, I–II, III–IV,
and V–VI), number of NFTs in CA1, and amyloid score (0 –3).
The three predictor variables were examined separately and
combined in a single analysis.
Results
General Characteristics of the Subjects
Twenty-eight cases and 56 control subjects were included
in the study. Age, gender, presence of cognitive deficits,
and the incidence of pathologically confirmed chronic
cardiac, respiratory, or neoplastic illnesses are summarized in Table 1. Of note is the higher percentage of severe
cardiac pathology (atherosclerotic disease) among the
control subjects. The prevalence of cancer and severe
respiratory illnesses was comparable in the two groups.
Three (10.7%) cases and seven (12.5%) control subjects
had clinical dementia.
Demographics of Suicide
Twelve of the 28 cases (43%) wrote a note of intent. Three
other individuals (11%) had a conversation with a relative
or friend that retrospectively could be interpreted as
manifesting suicidal intent. The specific suicide method is
shown in Table 2. The two most common methods were
firearms, utilized by eight individuals (28.6%), seven of
whom were men, and drug overdose, employed by eight
subjects as well, seven of whom were women.
Twenty-two cases had a psychologic autopsy performed. Analysis of the Axis I evaluation is shown in
Table 3. Sixteen cases (73%) had a diagnosable affective
illness; five of the 16 had psychotic features. Eleven (50%)
were in their first episode of major unipolar depression,
Ratios are male:female.
One case combined drug overdose and plastic bag asphyxia.
three had a recurrent major depressive episode, and two
had dysthymia. Five subjects had an alcohol or drug use
disorder, and four had other diagnosable conditions
(schizophrenia [n 5 1], hypochondriasis [n 5 1], and
undifferentiated somatiform disorder [n 5 2]). A clinical
diagnosis of dementia was made in three subjects.
Brain Weight
Brain weight did not differ between cases and control
subjects, but the range was broader among the latter
(Figure 1A). Possible explanations for this broader range
are the higher incidence of brain edema among subjects
dying of chronic illnesses (responsible for the upper brain
weight value). Looking at the lower range, none of the
cases but three of the control subjects had brains weighing
less than 1000 g. The three were women with clinical
dementia. The neuropathologic substrate of the dementia
was AD in one case, Huntington’s disease (and mild AD
pathology) in another, and multiple brain infarcts (and
minimal AD pathology) in the third. As would be expected
and can be seen in Figure 1B, brain weight declined with
age (p 5 .035), with a gender difference (p , .0001).
Brain weight decreased as modified Braak stage score
increased (Figure 1C). The lack of apparent brain weight
decline among subjects within the most severe modified
Braak stage is most likely due to the increased representation of men (with their inherent heavier brains) in that
group (eight of 10 subjects were male).
Microscopic Studies
Figure 2 shows the distribution of cases and control
subjects among the four modified Braak score groups.
Although a similar proportion of both cases and control
subjects fell within the mild AD pathology (stages I–II),
29% of the cases, as compared with 4% of the control
subjects, were in the highest severity stages (V–VI).
Concomitantly, there were fewer cases than control subjects (4% vs. 20%) with no neurofibrillary pathology
(stage 0). The number of NFTs in CA1 was higher for
cases than for control subjects (Figure 3A), and the values
140
A. Rubio et al
BIOL PSYCHIATRY
2001;49:137–145
Table 3. Axis I Diagnosis (DSM-III-R) in 22 Cases with Psychological Autopsy
1
Dementia
Major depression, single
episode
Major depression, recurrent
Dysthymia
Accompanying psychosis
Alcohol dependence
Alcohol abuse
Drug abuse
Schizophrenia
Hypochondriasis
Undifferentiated somatiform
disorder
a
2
3
4
5
6
7
8
9
10
ea
f
f
f
f
f
f
f
11
f
Case
12
f
f
f
14
15
16
17
18
19
20
21
22
f
f
f
f
f
f
f
13
f
f
f
f
f
f
f
f
f
f
f
f
f
f
f
Dementia of Parkinson’s disease.
increase proportionally according to the modified Braak
stage (Figure 3B). The stage of amyloid deposition in the
cortex was comparable for the two groups (Table 4). Five
cases and 12 control subjects displayed amyloid angiopathy. One case and two control subjects had Lewy bodies in
the parahippocampal gyrus, and ubiquitin-positive granular neuropil staining in the superficial layers of the
parahippocampal gyrus was identified in 10 (36%) cases
and 20 (36%) control subjects.
Regression Analysis
The results of both the separate and combined logistic
regression analyses were completely consistent. A higher
Braak score predicted case status either alone (p 5 .0028)
or combined with the other two predictor variables (p 5
.014). This difference can be seen by comparing the
frequency distributions for the Braak scores in the two
groups (Figure 2). The cases clearly have relatively more
scores of 3 (Braak stages V–VI) and relatively fewer
scores of both 2 (Braak stages III–IV) and 0 (no neurofibrillary pathology). There was no evidence for an association between amyloid score and case/control status either
alone (p 5 .71) or combined with the other two variables
(p 5 .75). This is consistent with the similarity of the
distribution of scores in the two groups. The evidence for
an association with NFTs was marginal (p 5 .16). The
median numbers of NFTs were 2 for cases and 1 for
control subjects; however, the 75th percentile was 10.5 for
cases and 2 for control subjects, indicating a much more
skewed distribution among the cases. This kind of difference suggests that the assumption of a linear relationship
on the logit scale that is required by the logistic regression
model may be questionable for this variable.
Discussion
Our results show increased AD pathology in a population
of elderly people committing suicide compared with a
group of age- and gender-matched control subjects who
died of natural causes. The prevalence of AD neuropathology in the control group is comparable to that seen in other
population studies when stratified by age (Braak and
Braak 1997). Due to the nature of the dependent variable
being tested, suicide, the study has some intrinsic limitations, which we tried to minimize in our design. Specifically we addressed the issue of selection and ascertainment biases as follows. The selection of cases was based
on age, history of successful suicide, and availability of
brain tissue. The clinical or neurologic status did not play
a role in the performance of the autopsy, acquisition of
brain tissue, or selection for the study. Because all suicides
receive a forensic autopsy, we do not believe that social,
educational, or other factors could have biased the selection. Choosing appropriate control subjects is probably the
key factor for any case– control study. Our purpose was to
select a representative sample of the general population,
comparable in age and gender to that of the cases. To
increase statistical power, we chose two control subjects to
match each case. We believe that in our community
subjects receiving an autopsy at the university hospital are
more representative of the general population as a whole
than those having necropsies at the medical examiner’s
office (violent deaths, unexpected and unexplained deaths,
or deaths of individuals lacking physician attendance). It is
probably true that individuals dying at the university
hospital are sicker than those committing suicide (see also
Table 2), given that they all die of natural causes. We do
not believe that this factor biased the results. In fact, one
of the genetic risk factors for cardiovascular disease (the
Suicide and Alzheimer’s in the Elderly
BIOL PSYCHIATRY
2001;49:137–145
141
Figure 2. Distribution of cases (top) and control subjects (bottom) by modified Braak stage.
Figure 1. Brain weight by case/control status (A), age and
gender (B), and modified Braak stage (C). Data in A and C are
represented by box plots. The upper and lower limits of a box
represent the 75th and 25th percentiles, respectively, and the line
within the box indicates median values. Whiskers are representative of the range.
presence of one or more apolipoprotein E allele 4) is a risk
factor for AD as well. If our control subjects had a higher
prevalence of ApoE4, we could expect a higher degree of
AD pathology in the control group, which was not the
case. Because the neuropathologic grading was done
blindly to the case status and to all clinical information, a
biased assessment cannot explain the difference in AD
pathology between cases and control subjects. The neuropathologic examination was limited because of the lack of
extensive brain tissue available for examination among the
cases and the need to use the same methodology for the
control subjects. Although we would prefer to have more
brain areas available for testing, we believe that our
modified scoring system is representative of the degree of
AD pathology in the entire brain. This belief is substantiated by a recent study (Harding et al 2000) where the
authors evaluated 72 brains in different stages of AD
pathology and concluded that semiquantitative evaluation
of NFTs in the hippocampus and inferior temporal cortices
resulted in AD scores comparable to those obtained by the
full original Braak method (k statistic of .97).
Another factor limiting the scope of this study is the
lack of homogeneity of clinical data for cases and control
subjects. Although medical records were reviewed in both
groups, cases had additional information resulting from
the psychologic autopsy. As is common in case– control
studies like this one, where death is the key factor for
subject inclusion, the level of neuropsychologic evaluation
142
A. Rubio et al
BIOL PSYCHIATRY
2001;49:137–145
Figure 3. The number of neurofibrillary tangles (NFTs) in CA1
of the hippocampus plotted by case/control status (A) and by
modified Braak stage (B). Data are represented by box plots. The
upper and lower limits of a box represent the 75th and 25th
percentiles, respectively, and the line within the box indicates
median values. Whiskers are representative of the range.
was quite variable among individuals. Because of this, we
did not include clinical status as a dependent variable and
limited our end points to objective neuropathologic data
obtained in a blind fashion. Thus, our results do not
attempt to link suicide with AD (in the clinico-pathologic
sense), but with Alzheimer’s neuropathology.
The cases included in our study shared the common
experience of planning and carrying out an emotionally
and intellectually challenging act. Postmortem psychologic evaluation was done in 22 of the 28 cases included in
this study. Three individuals were identified as having
dementia. A 79-year-old man had dementia of Parkinson
disease. His neuropathologic evaluation revealed Dementia with Lewy bodies and mild AD changes. Secondly, a
Table 4. Amyloid Deposition in the Parahippocampal Cortex
No amyloid plaques
Mild amyloid deposits
Moderate amyloid deposits
Severe amyloid deposits
Cases [n (%)]
Control subjects
[n (%)]
16 (57)
6 (21)
5 (18)
1 (4)
26 (46)
14 (25)
8 (14)
8 (14)
75-year-old man with chronic schizophrenia was also
diagnosed as having dementia. His neuropathologic examination revealed mild AD pathology (Braak stages I–II),
hardly justifying this patient’s cognitive decline. Lastly,
the brain of a minimally demented 92-year-old female
nursing home resident had severe AD pathology (Braak
stages V–VI). Lack of clinical dementia in individuals
with severe AD pathology has been reported (Davis et al
1999). Alzheimer’s disease is a clinico-pathologic entity
(Berg and Morris 1994), and the currently recommended
algorithm for the pathologic confirmation of AD requires
the presence of clinical dementia (The National Institute
on Aging and Reagan Institute Working Group on Diagnostic Criteria for the Neuropathological Assessment of
Alzheimer’s Disease 1997). We interpret this apparent
lack of clinico-pathologic correlation as follows. Incipient
Alzheimer’s pathology is the rule rather than the exception
in the brains of elderly individuals. There is consensus
among experts that these pathologic changes are progressive, and that the progression advances in a stereotyped
manner according to brain topography. The tempo of the
progression is slow, with the disease taking years to
decades to advance from one Braak stage to another. The
tempo would vary among different subjects, depending on
known and unknown risk factors, both genetic (Arendt et
al 1997; Li et al 1997; Nitsch et al 1996) and environmental (Butler et al 1996; Itzhaki et al 1997). On the other
hand, individual susceptibility (brain reserve) plays a role
in the personal threshold for developing clinical dementia.
Because dementia is the key for the clinical diagnosis of
AD, we frequently observe neuropathologically comparable individuals with regard to Alzheimer’s pathology in
very different cognitive stages. Thus, although the majority of humans would probably develop in time enough
Alzheimer’s pathology to manifest clinical dementia, for
those in whom neuropathology starts late in life and
progresses slowly, or who have a larger “brain reserve,”
AD, in the clinico-pathologic sense, will not be manifest
within a normal life span. On the other hand, although
dementia is a cardinal feature of AD, it is not the only one,
and surely not the first to occur. Emotional-affective
symptoms most likely precede dementia, and because
cognitive decline is necessary for the clinical diagnosis of
AD, emotional symptoms due to Alzheimer’s pathology
would not be recognized as AD. In a recently published
study (Smith et al 2000) of individuals carrying the
Huntington disease mutation (so that the disease is genetically determined, provided patients live long enough), the
authors were able to detect subtle movement abnormalities
predicting disease progression. Similarly, in the limited
fashion allowed by the case– control method, we suggest
that suicide may be a manifestation of AD pathology.
Indeed, the population of patients at risk for suicide
Suicide and Alzheimer’s in the Elderly
because of such emotional changes may be larger than the
population who are ever diagnosed with AD.
In contrast to the low prevalence of clinical dementia
among the cases, postmortem psychological autopsy conferences demonstrated that the majority of individuals
studied (79%) had some type of depressive disorder (Table
3), most commonly single-episode major depression (50%
of the subjects with postmortem psychologic evaluation).
Neuropathologic examination for the presence of Alzheimer’s-related changes in those subjects with a single
episode of major depression revealed that nine of 11
individuals had early (Braak stages I–II) Alzheimer’s
pathology. Depressive symptoms in these patients with
early Alzheimer’s changes could either be a reaction to the
subjects’ awareness of their cognitive decline or a direct
manifestation of the entorhinal/limbic degeneration of
incipient (preclinical) AD. Alternatively, it could be argued that neurochemical imbalances usually seen in depressed patients committing suicide (for a review, see
Mann 1998) may stress neuronal systems, inducing neurodegeneration by decreasing plasticity and synaptic repair
(Brezun and Daszuta 1999). In fact, reduced activity of
5-hydroxytryptamine (5-HT) has been detected in AD
patients (Meltzer et al 1998; Terry et al 1991) and has been
linked to emotional imbalances, depression, and psychosis
(Zubenko et al 1991; Zweig et al 1988). These neurochemical imbalances may predispose patients to aggressive or
impulsive acts such as suicide, just as altered indices of
central serotonin function have been associated with suicide (Ågren 1980, 1983; Bachus et al 1997; Banki et al
1984; Brown et al 1982; Coccaro et al 1990; Mann and
Malone 1997; Mann et al 1996; Montgomery 1987; Ninan
et al 1984; Roy et al 1986; Stockmeier and Meltzer 1995;
Tra¨skman et al 1981; van Praag 1983; Virkkunen et al
1989) and other violent, aggressive behaviors (Virkkunen
et al 1987, 1995).
The majority of serotonin-containing neurons are located in the brain stem, predominantly in the dorsal and
ventral raphe (Frazer and Hensler 1999). The low serotonergic status seen in suicide victims does not seem to be
due to fewer neurons in the raphe nuclei (Underwood et al
1999). On the other hand, medial raphe neurons, including
serotonergic neurons, have NFTs in AD (Halliday et al
1992). Axons from the median raphe neurons project to
the hippocampus and cerebral cortex, whereas from the
dorsal raphe they reach the striatum and cortex. Numerous
pre- and postsynaptic serotonin receptors have been identified and their genes mapped. Polymorphisms for 5-HT2A
and 5-HT2C may be responsible not only for susceptibility
to psychosis (Gutierrez et al 1996; Williams et al 1997),
but also for the psychotic symptomatology in patients with
AD (Holmes et al 1998). These findings may be gender
dependent, as indicated in animal models by the influence
BIOL PSYCHIATRY
2001;49:137–145
143
of aromatase conversion of androgens to estrogens in the
expression of 5-HT2A and serotonin transporters in the
brain (Fink et al 1999).
A third possible explanation of the association that we
have observed is that both suicidal behavior and Alzheimer’s pathology may be independently linked to a genetic
or environmental factor that affects both (Li et al 1997;
Oliveira et al 1998).
Our conclusion that people with moderate to severe AD
pathology are overrepresented in a sample of elderly
patients who completed suicide should be validated in a
larger, prospective study with psychologic autopsy data
for control subjects as well as for cases. The psychologic
examination should include cognitive, emotional, and
behavioral assessment. We hypothesize that the emotional-affective state of cognitively intact or minimally impaired individuals who later show AD symptoms may put
them at risk for suicide before they reach clinically
diagnosable AD.
Supported in part by grants AG-08665 and RO1 MH 54682.
The authors thank Ms. Frances Vito and Mr. Matt Frank for the
histochemical and immunohistochemical preparations.
Presented at the “Unique Features of Neuropsychiatric Disease in the
Elderly” panel, annual meeting of the American College of Neuropsychopharmacology, December 12–16, 1999, Acapulco, Mexico.
References
Aarsland D, Cummings JL, Yenner G, Miller B (1996): Relationship of aggressive behavior to other neuropsychiatric
symptoms in patients with Alzheimer’s disease. Am J Psychiatry 153:243–247.
Ågren H (1980): Symptom patterns in unipolar and bipolar
depression correlating with monoamine metabolites in the
cerebrospinal fluid: I. General patterns. Psychiatry Res
3:211–223.
Ågren H (1983): Life at risk: Markers of suicidality in depression. Psychiatry Dev 1:87–103.
Arendt T, Schindler C, Bruckner MK, Eschrich K, Bigl V,
Zedlick D, Marcova L (1997): Plastic neuronal remodeling is
impaired in patients with Alzheimer’s disease carrying apolipoprotein epsilon 4 allele. J Neurosci 17:516 –529.
Bachus SE, Hyde TM, Akil M, Shannon Weikert C, Vawter MP,
Kleinman JE (1997): Neuropathology of suicide: A review
and an approach. Ann N Y Acad Sci 836:201–219.
Bancher C, Jellinger K, Lassmann H, Fischer P, Leblhuber F
(1996): Correlations between mental state and quantitative
neuropathology in the Vienna Longitudinal Study of Dementia. Eur J Psychiatry Clin Neurosci 246:137–146.
Banki CM, Arato M, Papp Z, Kurcz M (1984): Biochemical
markers in suicidal patients. Investigations with cerebrospinal
fluid amine metabolites and neuroendocrine tests. J Affect
Disord 6:341–350.
Berg L, Morris JC (1994): Diagnosis. In: Terry RD, Katzman R,
144
BIOL PSYCHIATRY
2001;49:137–145
Bick KL, editors. Alzheimer’s Disease. New York: Raven,
9 –25.
Blumenthal SJ (1988): Suicide: A guide to risk factors, assessment and treatment of suicidal patients. Med Clin North Am
72:937–971.
Braak H, Braak E (1991): Neuropathological staging of Alzheimer-related changes. Acta Neuropathol 82:239 –259.
Braak H, Braak E (1997): Frequency of stages of Alzheimerrelated lesions in different age categories. Neurobiol Aging
18:351–357.
Braak H, Duyckaerts C, Braak E, Piette F (1993): Neuropathological staging of Alzheimer-related changes correlates with
psychometrically assessed intellectual status. In: Corain B,
Iqbal K, Nicolini M, Wiblad B, Wisniewski H, Zatta P,
editors. Alzheimer’s Disease: Advances in Clinical and Basic
Research. Chichester, UK: Wiley, 131–137.
Brezun JM, Daszuta A (1999): Depletion in serotonin decreases
neurogenesis in the dentate gyrus and the subventricular zone
of adult rats. Neuroscience 89:999 –1002.
Brown GL, Ebert MH, Goyer PF, Jimersom DC, Klein WJ,
Bunney WE, et al (1982): Aggression, suicide and serotonin:
Relationship to CSF amine metabolites. Am J Psychiatry
139:741–746.
Butler SM, Ashford JW, Snowdon DA (1996): Age, education,
and changes in the Mini-Mental State Exam scores of older
women: Findings from the Nun Study. J Am Geriatr Soc
44:675– 681.
Chiu HF, Lam LC, Pang AH, Leung CM, Wong CK (1996):
Attempted suicide by Chinese elderly in Hong Kong. Gen
Hosp Psychiatry 18:444 – 447.
Coccaro EF, Silver LJ, Owen K, Davis KL (1990): Serotonin in
mood and personality disorders. In: Coccaro EF, Murphy DL,
editors. Serotonin in Major Psychiatric Disorders. Washington, DC: APA Press, 71–97.
Conwell Y (1995): Dementia. Crisis 16:5– 6.
Conwell Y, Duberstein PR, Cox C, Herrmann JH, Forbes NT,
Caine ED (1996): Relationships of age and axis I diagnoses in
victims of completed suicide: A psychological autopsy study.
Am J Psychiatry 153:1001–1008.
Conwell Y, Rotenberg M, Caine ED (1990): Completed suicide
at age 50 and over. J Am Geriatr Soc 38:640 – 644.
Davis DG, Schnitt FA, Wekstein DR, Markesbery WR (1999):
Alzheimer neuropathologic alterations in aged cognitively
normal subjects. J Neuropathol Exp Neurol 58:376 –388.
Dickson DW (1997): The value of cross sectional neuroanatomical studies as a conceptual framework for prospective clinicopathological studies. Neurobiol Aging 18:382–386.
Draper B, Moore CM, Brodaty H (1998): Suicidal ideation and
the “wish to die” in dementia patients: The role of depression.
Age Ageing 27:503–507.
Eastley R, Wilcock GK (1997): Prevalence and correlates of
aggressive behaviours occurring in patients with Alzheimer’s
disease. Int J Geriatr Psychiatry 12:484 – 487.
Ferris SH, Hofeldt GT, Carbone G, Masciandaro P, Troetel WM,
Imbimbo BP (1999): Suicide in two patients with a diagnosis
of probable Alzheimer disease. Alzheimer Dis Assoc Disord
13:88 –90.
Fewster PH, Griffin-Brooks S, MacGregor J, Ojalvo-Rose E,
A. Rubio et al
Ball MJ (1991): A topographical pathway by which histopathological lesions disseminate through the brain of patients with Alzheimer’s disease. Dementia 2:121–132.
Fink G, Sumner B, Rosie R, Wilson H, McQueen J (1999):
Androgen actions on central serotonin neurotransmission:
relevance for mood, mental state and memory. Behav Brain
Res 105:53– 68.
Frazer A, Hensler JG (1999): Serotonin. In: Siegel GJ, editor.
Basic Neurochemistry. Molecular, Cellular and Medical
Aspects, 6th ed. Philadelphia: Lippincott, Williams and
Wilkins, 263–292.
Frierson RL (1991): Suicide attempts by the old and the very old.
Arch Intern Med 151:141–144.
Geddes JW, Snowdon DA, Soultanian NS, Tekirian TL, Riley
KP, Ashford JW, et al (1996): Braak stages III-IV of
Alzheimer-related neuropathology are associated with mild
memory loss, stages V-VI are associated with dementia:
Findings from the Nun study. J Neuropathol Exp Neurol
55:617.
Geddes JW, Tekirian TL, Soultanian NS, Ashford JW, Davis
DG, Markesbery WR (1997): Comparison of neuropathologic
criteria for the diagnosis of Alzheimer’s disease. Neurobiol
Aging 18(suppl 4):S99 –S105.
Gertz HJ, Xuereb JH, Huppert FA, Brayne C, Kruger H, McGee
MA, et al (1996): The relationship between clinical dementia
and neuropathological staging (Braak) in a very elderly
community sample. Eur Arch Psychiatry Clin Neurosci 246:
132–136.
Gutierrez B, Fananas L, Arranz MJ, Valles V, Guillamat R, van
Os J, Collier D (1996): Allelic association analysis of the
5-HT2C receptor gene in bipolar disorder. Neurosci Lett
212:65– 67.
Halliday GM, McCann HL, Pamphlett R, Brooks WS, Creasey
H, McCusker E, et al (1992): Brain stem serotonin-synthesizing neurons in Alzheimer’s disease: A clinicopathological
correlation. Acta Neuropathol 84:638 – 650.
Harding AJ, Kril JJ, Halliday GM (2000): Practical measures to
simplify the Braak tangle staging method for routine pathological screening. Acta Neuropathol 99:199 –208.
Harris EC, Barraclough B (1997): Suicide as an outcome for
mental disorders: A meta-analysis. Br J Psychiatry 170:205–
228.
Hepple J, Quinton C (1997): One hundred cases of attempted
suicide in the elderly. Br J Psychiatry 171:42– 46.
Holmes C, Arranz MJ, Powell JF, Collier DA, Lovestone S
(1998): 5-HT2A and 5-HT2C receptor polymorphisms and
psychopathology in late onset Alzheimer’s disease. Hum Mol
Genet 7:1507–1509.
Iqbal K, Braak H, Braak E, Grundke-Iqbal I (1993): Silver
labeling of Alzheimer neurofibrillary changes and brain b
amyloid. J Histotechnol 16:335–342.
Itzhaki RF, Lin WR, Shang D, Wilcock GK, Faragher B,
Jamieson GA (1997): Herpes simplex virus type 1 in brain
and risk of Alzheimer’s disease. Lancet 349:241–244.
Jellinger KA, Bancher C (1997): Proposals for re-evaluation of
current autopsy criteria for the diagnosis of Alzheimer’s
disease. Neurobiol Aging 18(suppl 4):S55–S65.
Kim KY, Hershey LA (1988): Diagnosis and treatment of
depression in the elderly. Int J Psychiatry Med 18:211–221.
Suicide and Alzheimer’s in the Elderly
Lecso PA (1989): Murder-suicide in Alzheimer’s disease. J Am
Geriatr Soc 37:167–168.
Li T, Homes C, Sham PC, Vallada H, Birkett J, Kirov G, et al
(1997): Allelic functional variation of serotonin transporter
expression is a susceoptibility factor for late onset Alzheimer’s disease. Neuroreport 8:683– 686.
Lyness JM, Conwell Y, Nelson JC (1992): Suicide attempts in
elderly psychiatric inpatients. J Am Geriatr Soc 40:320 –324.
Mann JJ (1998): The neurobiology of suicide. Nat Med 1998;4:
25–30.
Mann JJ, Malone KM (1997): Cerebrospinal fluid amines and
higher lethality suicide attempts in depressed inpatients. Biol
Psychiatry 41:162–171.
Mann JJ, Underwood MD, Arango V (1996): Postmortem studies
of suicide victims. In: Watson SJ, editor. Biology of Schizophrenia and Affective Disease, 1st ed. Washington, DC: APP,
197–220.
Margo GM, Finkel JA (1990): Early dementia as a risk factor for
suicide. Hosp Community Psychiatry 41:676 – 678.
Masliah E (1995): Natural evolution of the neurodegenerative
alterations in Alzheimer’s disease. Neurobiol Aging 16:280 –
282.
McCullagh P, Nelder JA (1989): Generalized Linear Models,
2nd ed. London: Chapman and Hall.
Meltzer CC, Smith G, DeKosky ST, Pollock BG, Mathis CA,
Moore RY, et al (1998): Serotonin in aging, late-life depression, and Alzheimer’s disease: The emerging role of functional imaging. Neuropsychopharmacology 18:407– 430.
Montgomery SA (1987): The psychopharmacology of borderline
personality disorders. Acta Psychiatr Belg 87:260 –266.
The National Institute on Aging, Reagan Institute Working
Group on Diagnostic Criteria for the Neuropathological
Assessment of Alzheimer’s Disease (1997): Consensus recommendation for the postmortem diagnosis of Alzheimer’s
disease. Neurobiol Aging 18(suppl 1):S1–S2.
Newell KL, Hyman BT, Growdon JH, Hedley-Whyte ET (1999):
Application of the National Institute on Aging (NIA)-Reagan
Institute criteria for the neuropathologic diagnosis of Alzheimer disease. J Neuropathol Exp Neurol 58:1147–1155.
Ninan PT, van Kammen DP, Scheinin M, Linnoila M, Bunney
WE, Goodwin FK (1984): CSF 5-hydroxyindoleacetic acid
levels in suicidal schizophrenic patients. Am J Psychiatry
141:566 –569.
Nitsch RM, Deng M, Growdon JH, Wurtman RJ (1996): Serotonin 5-HT2a and 5-HT2c receptors stimulate amyloid precursor protein ectodomain secretion. J Biol Chem 271:4188 –
4194.
Oliveira JR, Gallindo RM, Maia LG, Brito-Marques PR, Otto
PA, Passos-Bueno MR, et al (1998): The short variant of the
polymorphism within the promoter region of the serotonin
transporter gene is a risk factor for late onset Alzheimer’s
disease. Mol Psychiatry 3:438 – 451.
Rodhe K, Peskind ER, Raskind MA (1995): Suicide in two
patients with Alzheimer’s disease. J Am Geriatr Soc 43:187–
189.
Roy A, Ågren H, Pickar D, Linnoila M, Doran A, Cutler N, et al
BIOL PSYCHIATRY
2001;49:137–145
145
(1986): Reduced CSF concentrations of homovanilic acid and
homovanilic acid to 5-hydroxyindoleacetic acid ratios in
depressed patients: relationship to suicidal behavior and
dexamethasone nonsuppression. Am J Psychiatry 143:1539 –
1545.
Smith MA, Brandt J, Shadmehr R (2000): Motor disorder in
Huntington’s disease begins as a dysfunction in error feedback control. Nature 403:544 –549.
Spitzer RL, Williams JBW, Gibbon M (1986): Structured Clinical Interview for DSM-IIIR (SCID). New York: New York
State Psychiatric Institute, Biometrics Research.
Stockmeier CA, Meltzer HY (1995): Biologic studies of suicide
in schizophrenia. Clin Neuropharmacol 18(3):S9 –S17.
Terry RD, Masliah E, Salmon DP, Butters N, DeTeresa R, Hill
R, et al (1991): Physical basis of cognitive alterations in
Alzheimer’s disease: Synapse loss is the major correlate of
cognitive impairment. Ann Neurol 30:572–580.
Tra¨skman L, Åsberg M, Bertilsson L, Sjo¨strand L (1981):
Monoamine metabolites in CSF and suicidal behavior. Arch
Gen Psychiatry 38:631– 636.
Underwood MD, Khaibulina AA, Ellis SP, Moran A, Rice PM,
Mann JJ, Arango V (1999): Morphometry of the dorsal raphe
nucleus serotonergic neurons in suicide victims. Biol Psychiatry 46:473– 483.
van Praag HM (1983): CSF 5-HIAA and suicide in nondepressed schizophrenics. Lancet ii:977–978.
Virkkunen M, De Jong J, Bartko J, Linnoila M (1989): Psychobiological concomitants of history of suicide attempts among
violent offenders and impulsive fire setters. Arch Gen Psychiatry 46:604 – 606.
Virkkunen M, Goldman D, Nielsen D, Linnoila M (1995): Low
brain serotonin turnover rate (low CSF 5-HIAA) and impulsive violence. J Psychiatr Neurosci 20:271–275.
Virkkunen M, Nuutila A, Goodwin FK, Linnoila M (1987):
Cerebrospinal fluid monoamine metabolite levels in male
arsonists. Arch Gen Psychiatry 44:241–247.
Williams J, McGuffin P, Nothen MM, Owen MJ, EMASS
Consortium (1997): Meta-analysis of association between the
5-HT2a receptor T102C polymorphism and schizophrenia.
Lancet 349:1221.
Willis RH (1987): Suicide risk in elderly persons: Diagnosis and
management. Mt Sinai J Med 54:14 –17.
Wragg RE, Jeste DV (1989): Overview of depression and
psychosis in Alzheimer’s disease. Am J Psychiatry 146:577–
587.
Xuereb JH, Gertz HJ, Huppert F, Brayne C, Wischik CM,
Mukaetoba-Ladinska E (1995): The application of Braak’s
staging model of Alzheimer-type pathology to neuropathological diagnosis of dementia. Neuropathol Appl Neurobiol
21:44.
Zubenko GS, Moosy J, Martinez AJ, Rao G, Claassen D, Rosen
J, Kopp U (1991): Neuropathologic and neurochemical correlates of psychosis in primary dementia. Arch Neurol 48:
619 – 624.
Zweig RM, Ross CA, Hedreen JC, Steele C, Cardillo JE,
Whitehouse PJ, et al (1988): The neuropathology of aminergic nuclei in Alzheimer’s disease. Ann Neurol 24:233–242.