Age and Ageing 1998; 27: 477-483
© 1998, Bntish Genatncs Society
Occupation type as a predictor of
cognitive decline and dementia in
old age
ANTHONY
F, JORM,
BRYAN RODGERS,
A. Scon
HENDERSON, AILSA
E.
KORTEN, PATRICIA
A.
JACOMB,
HELEN CHRISTENSEN, ANDREW MACKINNON I
NH and MRC Psychiatric Epidemiology Research Centre, The Australian National University, Canberra, ACT 0200,
Australia
I The Mental Health Research Institute of Victoria, Parkville, Victoria 3052, Australia
Address correspondence to: A. Jorm, Fax: (+61) 6 249 0733. E-mail: Anthony.jorm@anu,edu,au
Abstract
Objective: to assess whether an individual's main occupation predicts cognitive decline or dementia.
Methods: the data were taken from a longitudinal study of 518 men aged 70 or over. Main occupation was coded
into one of John Holland's six occupational categories. The subjects completed four cognitive tests and were
diagnosed for dementia on two occasions three and a half years apart. The cognitive tests were the Mini-Mental
State Examination, Episodic Memory Test, Symbol-Letter Modalities Test and National Adult Reading Test.
Informants also completed the Informant Questionnaire on Cognitive Decline in the Elderly. Dementia was
diagnosed by the American Psychiatric Association Diagnostic and Statistical Manual of Mental Disorders, 3rd
edition, revised (DSM-III-R) and ICD-lO criteria using the Canberra Interview for the Elderly.
Results: cross-sectional analysis of the wave 1 data showed that the realistic occupations, which include trade,
technical and some service occupations, had poorer cognitive performance and a higher prevalence of DSM-III-R
dementia, These differences held even when age, education and native English were statistically controlled. The
greatest occupational difference was on the National Adult Reading Test, which estimates pre-morbid ability. By
contrast, there were no occupational differences in longitudinal change in cognitive test performance, informant
reports of cognitive decline or incident cases of dementia over three and a half years.
Conclusion: cross-sectional occupational differences on cognitive tests and in dementia prevalence are due to
differences in pre-morbid ability rather than to differences in rate of cognitive decline.
Keywords: cognitive tests, dementia, occupation
Introduction
Two recent cross-sectional studies have shown that
cognitive impairment in old age, as measured by the
Mini-Mental State Examination (MMSE) [1], varies
across occupational groups.
A French study classified principal life-time occupation into 23 categories and found that farm workers,
farm managers, domestic service employees and bluecollar workers had more cognitive impairment than
workers in intellectual occupations, even after controlling for age, sex, education, sensory impairment,
alcohol use and psychotropic drug use [2]. An Italian
study classified principal life-time occupation into six
categories and found that farmers scored lower on the
MMSE than white-collar workers [3]. This difference
held when age, education and financial dissatisfaction
were controlled. The authors considered several
possible explanations for these differences, including
occupational exposure to toxic chemicals, lower
cognitive stimulation throughout life, poorer medical
care and health behaviour and occupational differences
in pre-morbid intelligence.
There have also been studies comparing dementia in
high- and low-status occupations. A lJS study found
that people who had been in low-status jobs had a
higher risk of dementia, even when education was
controlled [4]. By contrast, a study from the UK found
no difference in risk of dementia between high- and
low-status occupations [5].
477
A. F. Jorm et 0/.
Studies of particular occupations have also been
used to test hypotheses that regular practice of a
cognitive skill retards age-related declines. For example, a study of spatial visualization ability found that
architects show as large an age-related decrement in
spatial visualization as unselected adults, despite
the constant use of spatial visualization skills in their
daily work [6]. Other studies of this type have also
suggested that age-related declines are still observed
when occupationally-relevant cognitive activities are
assessed [7].
This literature on the effects of occupational practice
can be seen as a subset of more general research on the
effects of mental or physical activity on cognitive
ageing. There is some evidence, from both psychological and physiological studies, that activity may have
some preventive effects on age-related cognitive
decline and dementia [8 -13]. However, in contrast to
the studies of life-time occupation, most of this
evidence comes from studies of activity during old
age rather than over the adult life-span.
While the existing evidence on the effects of lifetime occupation is suggestive, it is largely based on
cross-sectional data on cognitive functioning. Longitudinal studies are needed to provide more convincing
evidence on the issue. Reported here are the results
from a three-and-a-half year longitudinal study of
elderly men, in which occupation was coded according
to the classification ofJohn Holland [14].
Holland's taxonomy provides a classification of
occupations according to their psychological demands
which is well-suited to a test of the hypothesis that lifetime occupational activity may affect the rate of
cognitive ageing. The taxonomy places occupations
into six categories: 'realistic' (includes skilled trades,
technical, some service occupations), 'investigative'
(includes scientific and some technical occupations),
'artistic' (includes artistic, musical and literary occupations), 'social' (includes educational and social welfare
occupations), 'enterprising' (includes managerial and
sales occupations) and 'conventional' (includes office
and clerical occupations). Holland's taxonomy has
been the subject of much research and is widely
accepted in vocational guidance.
Occupational groups tend to differ in their level of
education as well as in the frequency of non-English
speaking background. A low level of education has
itself been suggested as a risk factor for dementia [15,
16]. To ensure that any occupational differences are
not due to education or native English, these factors
were statistically controlled.
Methods
or over from Canberra and Queanbeyan. Subjects were
selected from the compulsory electoral roll and from a
census of people in residential care. The wave 1
interviews of the subjects were undertaken in 19901991 and the wave 2 interviews in 1994. The average
time between interviews was 3.6 years. At both waves,
an interview was carried out with an informant (where
available) as well as the subject. Only the data from
male subjects were used here because many of the
women had spent much of their lives in household
duties.
At wave 1, there were 531 male subjects (representing 76% of those approached), 480 living at home and
51 in residential care. Of these, 518 subjects had
sufficient occupational data to code and could be
classified into a single occupational category. Three
hundred and eighty-one of these subjects had both
subject and informant interviews, 106 had only subject
interviews and 31 had only informant interviews. Wave
2 data were available for 329 of these subjects since
141 subjects had died and 48 refused or could not be
contacted. Of the 329 subjects at wave 2, there were
272 who had both subject and informant interviews,
28 who had only subject interviews and 29 who had
only informant interviews.
Interview content
Interviews were carried out by professional social
survey interviewers. The interviews incorporated the
Canberra Interview for the Elderly which provides
diagnoses of dementia and depression by both American Psychiatric Association Diagnostic and Statistical
Manual of Mental Disorders, 3rd edition, revised (DSMIII-R) and International Classification of Diseases, 10th
edition (ICD-IO) criteria [17, 18], as well as the
following cognitive tests: MMSE [1], the National
Adult Reading Test (NARn [19]' the Symbol-Letter
Modalities Test [20], the Episodic Memory Test: [21]
and the Informant Questionnaire on Cognitive Decline
in the Elderly (IQCODE) [22-23].
Coding of occupations
At the wave 1 interview, subjects were asked what was
their main job during their working life. Coding was
carried out using a list of 920 occupations from the
Australian Standard Classification of Occupations
which have been assigned to the John Holland
categories based on the judgements of an expert
panel [24]. Coding was carried out by a single person,
but independent coding of data from 30 subjects gave
97% agreement.
Socio-demographic characteristics
Sample
The subjects were a sample of elderly people aged 70
478
Questions to the subjects on education were used to
derive a measure of years of education. Subjects were
Occupation type, cognitive decline and dementia
also asked their native language (which was English for
88%).
Statistical analysis
Data analysis was both cross-sectional (wave 1) and
longitudinal.
In the cross-sectional analysis, the occupational
groups were compared for mean scores on the cognitive
tests using one-way analysis of variance and for
differences in dementia prevalence using Fisher's
exact test. To see whether any occupational group
differences remained once socio-demographic factors
were controlled, hierarchical multiple linear regression
analysis was used, with the cognitive test scores at wave
1 as the dependent variables. Age was entered as a
predictor on the first step, years of education and
English language on the second step and occupation on
the final step. Occupation was coded as a set of dummy
variables with realistic occupations as the reference
category. The realistic occupations were chosen as the
reference category because they were the largest group
and because one-way analyses of variance showed them
to have poorer test scores. The R2 change for the
regreSSion model was used as an index of the association
of each set of predictor variables with the wave 1 test
score, controlling for the effects of the predictor
variables added earlier. Standardized (3 weights were
used to assess the contribution of each occupational
category to the overall effect of occupation. Similar
hierarchical logistic regression analyses were carried out
with dementia diagnosis as the dependent variable.
In the longitudinal analysis, the aim was to assess
cognitive decline in previously non-demented individuals. Therefore, subjects who were demented at wave
1 were excluded. In studying change, two alternative
approaches are to analyse the difference between wave
1 and wave 2 scores (the difference score approach) or
to treat wave 1 scores as a covariate in regression
analysis where wave 2 score is the dependent variable
(the conditional regression approach). Both of these
approaches have strengths and weaknesses, and both
were used on the present data. However, the difference
score approach proved to be preferable because of the
operation of Lord's paradox with the conditional
regression approach in the present data [25]. (Lord's
paradox occurs in conditional regression analysis
where there are pre-test differences in group means
and scores at post-test regress to their respective group
means.) The dependent variables were change on the
cognitive tests from wave 1 to wave 2 and incidence of
new dementia cases at wave 2. Means were compared
using one-way analysis of variance and incidence rates
using Fisher's exact test.
Results
The number of subjects in each occupational category
was: artistic 9, conventional 51, enterprising 120,
investigative 60, realistic 216 and social 62. Thirteen
subjects could not be assigned to an occupational
category.
Table 1 shows the socio-demographic characteristics
of the occupational groups at wave 1. There was no
difference in age, but there were differences in years of
education and the percentage whose native language
was English.
Table 1 also shows the mean test scores and
dementia prevalence in the occupational groups.
There were differences on all the cognitive tests, but
not on the IQCODE. There were also differences in
DSM-III-R dementia. The lowest mean scores and the
highest prevalence of dementia were found in the
realistic occupations. However, the realistic group also
had the lowest level of education and a lower
percentage of native English speakers.
To control for the effects of age, education and native
English, a series of hierarchical multiple regressions
were carried out, with the wave 1 test scores as the
dependent variables. As shown in Table 2, there were
still occupational differences when these other factors
were controlled. The largest effect of occupation was
found with the NART (9% increment to R2) and the
smallest with the MMSE (2% increment to R 2 ). The (3
values associated with occupational groups showed
that the realistic group performed worse than each of
the others. The results of the regression analysis with
the IQCODE as the dependent variable are not shown
in Table 2, but there was again no effect of occupation
when the other predictors were controlled.
Hierarchical logistic regressions were also carried
out with dementia diagnoses as the dependent
variables. With DSM-III-R dementia as the dependent
variable, occupation added Significantly to the model
2
(x for improvement = 12.2, dJ = 5, P= 0.03). However, none of the (3 values for individual occupations
was Significant. With ICD-IO dementia as the dependent variable, occupation did not add significantly to
the model (P=0.38).
Based on the longitudinal data, Table 3 shows the
mean change scores on the cognitive tests for each
of the occupational groups. One-way analyses of
variance showed no differences between the groups.
Hierarchical regressions in which age, education and
native English were controlled did not alter these
findings.
When incident cases of dementia were examined,
the numbers were quite small (13 cases of DSM-III-R
dementia and six of ICD-I0). Because of the small
number of cases in each occupational group, the
groups were collapsed into two: realistic occupations
and other occupations. The incidence of DSM-III-R
dementia was non-significantly higher in the realistic
group (7.0% versus 3.9%, P= 0.26), as was the
incidence of ICD-lO dementia (4.5% lJersus 1.1%,
p=o.mn.
479
~
co
e
0.0
0.0
(9.0)
(2.2)
(16.3)
(11.9)
(0.1)
Conventional
8.7
2.1
26.5
10.4
99.6
116.4
3.2
(4.7)
(3.3)
(15.5)
(7.8)
(0.3)
78.1 (6.4)
12.2 (2.6)
89.8
(n = 51)
6.8
1.8
27.6
11.0
97.8
114.1
3.2
(2.8)
(3.4)
(17.4)
(9.0)
(0.4)
76.4 (5.3)
11.9 (2.5)
92.9
Enterprising
= 120)
(n
Investigative
0.0
0.0
28.0
11.5
101.5
116.3
3.2
(1.6)
(3.1)
(14.3)
a.8)
(0.4)
77.6 (5.9)
13.5 (2.8)
89.5
(n = 60)
Realistic
13.2
4.5
25.6
9.4
85.9
105.7
3.2
(4.2)
(3.5)
(17.6)
(9.4)
(0.6)
77.1 (5.3)
10.2 (2.4)
78.1
(n = 216)
Social
1.8
1.7
28.0
11.9
104.4
117.9
3.1
(2.6)
(2.3)
(16.3)
(8.1)
(0.3)
76.0 (5.0)
13.9 (2.8)
88.5
(n = 62)
Significant difference between occupational groups, ap<0.05; b p < 0.01.
MMSE, Mini-Mental State Examination; NART, National Adult Reading Test; IQCODE, Informant Questionnaire on Cognitive Decline in the Elderly; DSM-I1I-R, American Psychiatric Association
Diagnostic and Statistical Manual of Mental Disorders, 3rd edition, revised; ICD-lO, International Classification of Diseases, 10th edition (World Health Organisation, 1993),
Dementia prevalence (%)
DSM-III·Ra
ICD-lO
24.9
12.2
102.8
111.4
3.1
77.7 (5.2)
11.6 (3.0)
55.6
Socio-demographic
Age [mean (SD)]
Education, years [mean (SD)]b
% native English speakersb
Test score [mean (SD)]
MMSEb
Episodic memoryb
Symbol-Ietterb
NARTb
IQCODE
Artistic
(n = 9)
Characteristic
Occupation group
Table I. Differences between occupational groups at wave 1
:-
Q
~
....
3
o""S
-
1>
:n
.110.
0)
Change
0.02
P-value
0.06
0.00
0.00
P-value
0.07
0.15
0.16
0.19
0.11
0.10
0.05
-0.33
Value
0.13
0.00
0.00
0.00
0.02
0.05
0.29
0.00
P-value
0.20
0.16
0.11
Value
Change
0.04
0.04
0.11
Change
P-value
0.00
0.00
0.00
P-value
0n
0.20
0.11
0.20
0.19
0.20
0.23
0.12
Type of occupation
Conventional
Enterprising
Investigative
Social
Artistic
-0.36
Age
Socio-demographic
Education
Native English
Value
Predictor variable
(3
0.00
0.00
0.00
0.00
0.00
0.00
0.01
0.00
P-value
0.33
0.26
0.14
Value
0.07
0.12
0.14
0.00
0.00
0.00
0.25
0.24
0.20
0.25
0.09
0.35
0.27
-0.07
Value
(3
0.00
0.00
0.00
0.00
0.02
0.05
0.29
0.06
P-value
0.38
0.29
0.01
Value
R2
National Adult Reading Test
0.09
0.29
0.01
0.00
0.00
0.07
~.
,.,.:::I
3
tD
tD
c.
c.
:::I
~
:::I
tD
n
tD
C.
tD
:C'
:::I
;::t."
OQ
0
"'tJ
tD
.n
"<
,.,.
"'tJ
o·~:::I
R2
0.22
0.10
0.11
Change
R2
Symbol-letter modalities
0.25
0.02
0.01
0.04
0.67
0.21
0.11
Value
(3
n
c
0.05
0.12
0.13
0.10
-0.02
Type of occupation
Conventional
Enterprising
Investigative
Social
Artistic
0.00
0.00
0.00
P-value
R2
Episodic Memory Test
Dependent variable
0.18
0.18
-0.32
Age
Socio-demographic
Education
Native English
Value
Predictor variable
(3
Mini-Mental State Examination
Dependent variable
Table 2. Hierarchical regressions predicting cognitive test scores at wave 1
A. F. Jorm et al.
Table 3. Change on cognitive tests between waves for the various occupational groups
Mean (and SD) change a , by occupational group
Artistic
= 9)
Test
(n
MMSE
Episodic memory
Symbol-letter
NART
-0.3
-0.3
-0.5
-3.0
(2.6)
(2.3)
(S.5)
(S.4)
Conventional
(n
= 51)
0.6 (2.0)
-0.2 (2.7)
1.601.7)
1.9 (5.2)
Enterprising
(n
= 120)
0.30.S)
-0.7 (2.3)
2.202.3)
-0.9 (5.0)
Investigative
(n
= 60)
0.60.4)
-0.9 (2.5)
3.3 (9.6)
-0.2 (4.6)
Realistic
(n
= 216)
1.1 (2.3)
-0.6 (2.5)
4.002.9)
-0.2 (5.6)
Social
(n
= 62)
0.2 (2.0)
-0.1 (2.4)
3.000.3)
-1.1 (3.1)
aWave I score-wave 2 score.
MMSE, Mini-Mental State Examination; NART, National Adult Reading Test.
There are no significant differences between groups, P< 0.05.
Discussion
In cross-sectional analyses, the lowest test performance
and highest prevalence of dementia were found in the
realistic occupations which involve skill trades, technical and some service occupations. This finding is
consistent with the earlier studies [2, 3]' because
occupations like farm worker, farm manager, domestic
service employee and blue-collar worker all belong to
the realistic category in Holland's classification. While
Holland's taxonomy is relatively crude compared with
a coding of specific occupations, it has the advantage of
focusing on the common psychological demands of
broad groups of occupations. The realistic occupations
involve greater use of manual skills and less use of
literacy and other intellectual skills.
When the pattern of cross-sectional deficits was
examined, the largest occupational difference was
found on the NART and the smallest on the MMSE.
No occupational differences were found on the
IQCODE, which measures cognitive change. This
pattern of deficits suggests that the occupational
differences are in pre-morbid ability rather than in
cognitive decline. The longitudinal data are consistent
with this interpretation, with no differences between
occupational groups in rate of cognitive decline.
When dementia was examined, there were crosssectional differences in the prevalence of DSM-III-R
dementia, but not lCD-to dementia. We have previously reported data that DSM-I1I-R dementia is
affected by education, whereas lCD-to dementia is
not [26]. The difference may come about because the
diagnostic criteria for DSM-III-R dementia require only
cognitive impairment, while the lCD-tO criteria require
cognitive decline [27, 2S]. No differences were found
between the realistic and other occupational groups in
incident cases of dementia at wave 2. However, the
number of cases was small and the power of the
statistical analysis consequently limited.
In conclusion, occupational differences can be
demonstrated in cross-sectional data, controlling for
education, but these differences disappear with longitudinal measures of cognitive decline. It seems to be
482
pre-morbid intelligence that accounts for the differences between occupational groups. These results fit
Mortimer's [29] hypothesis that psychosocial risk
factors for dementia "act primarily to increase vulnerability, to reduce the margin of intellectual reserve to a
level where a more modest level of brain pathology
results in diagnosable dementia".
Acknowledgements
This study was supported by a grant from the
Australian Rotary Health Research Fund. We wish to
thank Ruth Scott, Colleen Doyle, Susan Lindsay,
Suzanne Dee, Karen Maxwell for assistance with the
study.
Key points
•
•
•
•
Elderly men who had worked in manual (,realistic')
occupations (which include trade, technical and
some service occupations) performed worse on
cognitive tests and had a higher prevalence of DSMIII-R dementia.
The poorer performance of the men in realistic
occupations was not entirely due to a lower level of
education.
When these men were followed over three and a
half years there were no differences between
occupational groups in rate of cognitive decline.
The lower cognitive performance of the men in
realistic occupations was due to pre-morbid ability
rather than a faster rate of cognitive decline in old
age.
References
1. Folstein MF, Folstein SE, McHugh PRo 'Mini-Mental State': a
practical method for grading the cognitive state of patients
for the clinician.] Psychiatr Res 1975; 12: 189-98.
2. Dartigues J-F, Gagnon M, Letenneur L et al. Principal
lifetime occupation and cognitive impairment in a French
elderly cohort (PAQUID). Am] Epidemioll992; 135: 981-8.
Occupation type, cognitive decline and dementia
3. Frisoni GB, Rozzini R, Bianchetti A, Trabucchi M. Principal
lifetime occupation and MMSE score in elderly persons.
J Gerontol Soc Sci 1993; 48: S31O-4.
4. Stern Y, Gurland B, Tatemichi TK, Tang MX, Wilder D,
Mayeux R. Influence of education and occupation on the
incidence of Alzheimer's disease. JAMA 1994; 271: 1004-10.
5. Paykel ES, Brayne C, Huppert FA et al. Incidence of
dementia in a population older than 75 years in the United
Kingdom. Arch Gen Psychiatry 1994; 51: 325-32.
6. Salthouse TA, Babcock RL, Skovronek E, Mitchell DRD,
Palmon R. Age and experience effects in spatial visualization.
Develop Psycholl990; 26: 128-36.
7. Salthouse TA. Influence of experience on age differences
in cognitive functioning. Hum Factors 1990; 32: 551-69.
8. Arbuckle TY, Gold DP, Andres D. Cognitive functioning of
older people in relation to social and personality variables.
Psychol Aging 1986; 1: 55-62.
9. Christensen H, Korten AE,Jorm AF, Henderson AS, Scott R,
Mackinnon AJ. Activity levels and cognitive functioning in an
elderly community sample. Age Ageing 1996; 25: 72-80.
10. Fabrigoule C, Letenneur L, Dartigues JF, Zarrouk M,
Commenges D, Barberger-Gateau P. Social and leisure
activities and risk of dementia: a prospective longitudinal
study. J Am Geriat Soc 1995; 43: 485-90.
11. Hultsch DF, Hammer M, Small BJ. Age differences in
cognitive performance in later life: Relationships to selfreported health and activity life style. J Gerontol Psychol Sci
1993; 48: PI-II.
in the diagnosis of dementia and depression. Acta Psychiatr
Scand 1993; 87: 146-51.
18. Social Psychiatry Research Unit. The Canberra interview
for the elderly: a new field instrument for the diagnosis of
dementia and depression by ICD-lO and DSM-III-R. Acta
Psychiatr Scand 1992; 85: 105-13.
19. Nelson HE. The National Adult Reading Test (NARD Test
Manual. Berkshire: NFER Nelson, 1982.
20. Christensen H, Jorm AF, Henderson AS, Mackinnon AJ,
Korten AE, Scott LR. The relationship between health and
cognitive functioning in an elderly community sample. Age
Ageing 1994; 23: 204-12.
21. Jorm AF. Use of informants' reports to study memory
changes in dementia. In: Rickman Led. Memory Functioning
in Dementia. Amsterdam: North-Holland, 1992.
22. Jorm AF, Jacomb PA. The informant questionnaire on
cognitive decline in the elderly (IQCODE): socio-demographic correlates, reliability, validity and some norms.
Psychol Med 1989; 19: 1015-22.
23. Jorm AF, Christensen H, Henderson AS, Jacomb PA,
Korten AE, Mackinnon A. Informant ratings of cognitive
decline of elderly people: relationship to longitudinal change
on cognitive tests. Age Ageing 1996; 25: 125-9.
24. Lokan]. The Self-directed Search: professional manual,
Australian supplement. Camberwell, Victoria: Australian
Council for Educational Research, 1988.
25. Plewis I. Analysing Change: Measurement and Explanation Using Longitudinal Data. Chichester: John Wiley, 1985.
12. Rogers RL, Meyer JS, Mortel KF. After reaching retirement age physical activity sustains cerebral perfusion and
cognition. J Am Geriat Soc 1990; 38: 123-8.
26. Henderson AS, Jorm AF, Mackinnon AJ et al. A survey
of dementia in the Canberra population: experience with
ICD-lO and DSM-III-R criteria. Psychol Med 1994; 24: 473-82.
13. Swaab DF. Brain aging and Alzheimer's disease, 'wear and
tear' versus 'use it orlose it'. NeurobiolAging 1991; 12: 317 -24.
27. American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders, third edition, revised.
Washington, DC: American Psychiatric Association, 1997.
14. Holland J1. Making Vocational Choices: a theory of
personalities and work environments. Englewood Cliffs, N]:
Prentice-Hall, 1985.
15. Mortimer JA, Graves AB. Education and other socioeconomic determinants of dementia and Alzheimer's disease.
Neurology 1993; 43 (suppl. 4): S39-44.
16. Katzman R. Education and the prevalence of dementia
and Alzheimer's disease. Neurology 1993; 43: 13-20.
17. Mackinnon A, Christensen H, Cullen JS et al. The
Canberra interview for the elderly: assessment of its validity
28. World Health Organization. The ICD-10 Classification of
Mental and Behavioural Disorders: diagnostic criteria for
research. Geneva: WHO, 1993.
29. Mortimer JA. Do psychosocial risk factors contribute to
Alzheimer's disease? In: Henderson AS, Henderson JH eds.
Aetiology of Dementia of Alzheimer's Type. Chichester: John
Wiley, 1988; 39-52.
Received 23 April 1997; accepted 23 May 1997
483
Latin America. © Help the Aged.