Content uploaded by David Gaist
Author content
All content in this area was uploaded by David Gaist on Feb 06, 2014
Content may be subject to copyright.
JOURNAL OF AGING AND HEALTH / February 2001
Nybo et al. / THE DANISH 1905 COHORT
The Danish 1905 Cohort:
A Genetic-Epidemiological Nationwide Survey
HANNE NYBO
DAVID GAIST
BERNARD JEUNE
LISE BATHUM
Odense University
MATT MCGUE
University of Minnesota
JAMES W. VAUPEL
Odense University; Max Planck Institute
for Demographic Research, Germany
KAARE CHRISTENSEN
Odense University
Objectives: The authors studied nonagenarians, a rapidly growing age group whose
cognitive and physical abilities have yet to be investigated systematically. Methods:
All Danes born in 1905 were invited to participate in a home-based 2-hour multidi-
mensional interview, including cognitive and physical performance tests and collec-
tion of DNA, carried out by lay interviewers. Population-based registers were used to
evaluate representativeness. Results: There were 2,262 participants. A total of 1,632
(72%) gave a DNA sample. Participants and nonparticipants were highly comparable
with regard to marital status, institutionalization, and hospitalization patterns, but
men and rural area residents were more likely to participate. Six months after the sur
-
vey began, 7.2% of the participants and 11.8% of the nonparticipants had died. Dis
-
cussion: Despite the known difficulties of conducting surveys among the extremely
old, it was possible to conduct a nationwide survey, including collection of DNA,
among more than 2,000 fairly nonselected nonagenarians using lay interviewers.
The Danish 1905 cohort study is a nationwide survey of all Danes born
in 1905. The overall goal is to establish a genetic-epidemiological
database to shed light on the aging process among the extremely old.
We wanted to describe the extremely old, focusing on physical and
32
JOURNAL OF AGING AND HEALTH, Vol. 13 No. 1, February 2001 32-46
© 2001 Sage Publications, Inc.
cognitive functioning, and to examine the feasibility of using lay inter
-
viewers to test physical functioning and to collect DNA.
Since the 1950s, the mortality rate among the oldest old in the
developed countries has declined dramatically (Vaupel et al., 1998),
causing a big rise in the number of oldest old (80+ years). Whereas both
centenarians and octogenarians are by now relatively well described,
only a limited number of smaller surveys focusing on nonagenari
-
ans has been conducted (Forsell, Jorm, von Strauss, & Winblad, 1995;
Harris, Finucane, Healy, & Bakarich, 1997; Heeren, Lagaay, von Beek,
Rooymans, & Hijmans, 1990; Meller, Fichter, Schroppel, & Beck
Eichinger, 1993; O’Connor, Pollitt, Brook, & Reiss, 1989; Ravaglia
et al., 1997). Although many of the cross-sectional and longitudinal
population-based studies of the elderly include nonagenarians, data
on this group are seldom presented separately, possibly due to the
small number participating. Furthermore, most of the ongoing longi-
tudinal surveys of younger elderly will not provide information on the
extremely old (90+ years) because the majority of the study popula-
tion will have deceased before they reach the extreme ages. The pres-
ent study has bypassed these problems by conducting a nationwide
survey among 2,262 nonagenarians. Five percent of the cohort is
expected to live to be 100 years old, which will provide sufficient
power to allow detailed studying of a large number of predictive fac-
tors for loss of abilities and mortality/survival in the follow-up sur-
veys. In this article, the method, study population, and feasibility of
the study are described.
Participants and Methods
The only inclusion criteria in the survey was that the participants
were born in 1905 and living in Denmark. All nonagenarians from this
birth cohort were approached irrespective of residence, health, and
cognitive status. Nonagenarians living in institutions were also
approached. There were no exclusion criteria.
Nybo et al. / THE DANISH 1905 COHORT 33
AUTHORS’ NOTE: This study and the activities of the Danish Center for Demographic Re
-
search are supported by a grant from the Danish National Research Foundation. Address reprint
request to Kaare Christensen, Epidemiology, Institute of Public Health, Sdr. Boulevard 23A.1,
DK-5000 Odense C.
The small size of the country (5.3 million inhabitants, 43,000 square
km) and the extensive registration of the inhabitants make Denmark
an attractive setting for population-based studies. All inhabitants have
a 10-digit unique and permanent personal identification number,
which includes the date of birth and gender of the person (ID number).
The ID numbers are recorded in the Danish Civil Registration System
(DCRS), along with demographic information, including date of death.
In early 1998, we retrieved data on ID numbers, names, addresses, and
marital status of all Danes living in Denmark who were born in 1905.
Hence, eligible participants were 92 or 93 years of age at the time of
the survey. The survey was approved by the Central Scientific Ethical
Committee of Denmark.
All persons in the cohort received a letter explaining the study and
asking permission for an interviewer to come to their residence to con-
duct a health-related, face-to-face interview and test their mental and
physical functioning. They were also asked to give a sample of cells
from which DNA could be isolated. If the person was unable to partic-
ipate due to physical or mental impairment, a proxy responder was
encouraged to participate in the interview instead. The nonagenarians
were contacted within 14 days by the interviewer, either by personal
contact or by phone, to obtain consent to participate in the survey.
The nonagenarians were considered nonparticipants if they did not
wish to participate in person or by proxy or if at least three attempts to
contact them, at different times, were unsuccessful.
During a 3-month period in 1998 (August to October), 93 inter
-
viewers from the Danish National Institute of Social Research carried
out the survey. The interviewers are not medically or paramedically
trained but do, however, have substantial experience in interviewing
the elderly (Christensen, Holm, McGue, Corder, & Vaupel, 1999). All
interviewers received a detailed training program by physicians just
before the start of the survey and were closely monitored during the
interview period.
A pilot study was performed in the spring of 1998, composed of a
total of 200 persons from the 1905 cohort. This study resulted only in
minor changes of the interview, and these data are included in the
overall analysis.
The interview part of the survey took, on average, 1 1/2 hours to
perform. Questions concerning the following topics were asked:
34 JOURNAL OF AGING AND HEALTH / February 2001
Sociodemographic factors: living conditions, educational attainment, and
work.
Anthropometric measures: self-reported height and weight (current, at
age 25, maximal weight ever).
Lifestyle habits: previous and current smoking and drinking behavior.
Health: self-rated health, diseases diagnosed by a physician, symptoms of
pulmonary and heart diseases, incontinence, pain, fall events, and frac
-
tures. All drugs, vitamins, and alternative medications taken on a regu
-
lar basis were recorded.
Physical ability: Avlund’s physical activity of daily living scale (Avlund
et al., 1991), Questions from Katz’s ADL scale (Katz et al., 1970) and
from Nagi’s measures of functional limitations (Nagi, 1976).
Psychological status: Symptoms of depression were measured by a modi
-
fied scale from the Cambridge Mental Disorders of the Elderly Exami
-
nation (McGue & Christensen, 1997; Roth et al., 1986). An informant
questionnaire (DECO) (Ritchie & Fuhrer, 1992) was used for evalua-
tion of cognitive impairment in the proxy interview.
Sensoric deficits: self-reported hearing and vision impairments.
Family history: parents’ age at death, participant’s age at birth of first and
last child.
Social life: contacts with friends and family, participation in social
activities.
The questionnaire, with minor changes, has previously been used in
the Longitudinal Survey of Aging Danish Twins (Christensen et al.,
1998; Christensen et al., 1999; McGue & Christensen, 1997).
Participants not able to participate in the face-to-face interview par
-
ticipated through a proxy responder; in most cases, this was a close
relative who went through a questionnaire identical to that used in the
face-to-face interviews.
Testing of cognitive and physical functioning lasted approximately
30 minutes. The following tests were included:
Cognitive abilities: Mini Mental State Examination (M. F. Folstein, S. E.
Folstein, & McHugh, 1975), verbal fluency, forward and backward
digit span, and immediate and delayed recall of 12 words.
Physical performance: ability to lift a 2.7 kg box above the head, flexibil
-
ity (ability to bring hands to neck, loin, and opposite toe), single chair
stand with or without the use of arms, and timed walk over a distance of
3 meters. These tests were, with minor changes, performed according
Nybo et al. / THE DANISH 1905 COHORT 35
to the protocol from the Women’s Health and Aging Study (Guralnik,
Fried, Simonsick, Kasper, & Lafferty, 1995). Handgrip was tested using
a handheld dynamometer (SMEDLEY’S, dynamometer TTM) for three
performances with the strongest hand. Pulmonary function was mea
-
sured using a small digital spirometer (MicroDL, Micro Medical Wtc)
among approximately half of the participants who were living in the
western part of the country. Because we did not have sufficient funds to
purchase digital spirometers for all 93 interviewers, participants living
in the eastern part of the country had their pulmonary function tested
using a peak-flow meter (Mini-Wright).
The DNA sample could either be given as a bloodspot or as a cheek
swab. The participants were asked to prick a finger with a special ster
-
ile automatically retractable lancet (Tenderfoot
Ò
ICT, USA). Blood-
spots were then collected on filter paper. Approximately 5 square cm
of the filter paper was soaked with blood. A bloodspot was defined as
having a poor quality if no more than 1 square cm of the filter paper
was filled with blood or not soaked through. Alternatively, DNA was
sampled by a swab from the inner side of the cheek. The swabs were
received in the laboratory within 24 to 48 hours after collection.
Participants and nonparticipants were compared with regard to sex,
marital status, type of housing, and area of residence (rural/urban)
using data from the DCRS. Data from this registry were also used to
estimate the mortality in the two groups 6 months after the start of the
survey.
Information on hospital discharges was available from a population-
based register with complete coverage of Funen County (approxi
-
mately 9% of the Danish population) for the period from 1973 to 1998.
We retrieved data on all 1905 cohort members residing in the county
(n = 398), and we calculated the total length of hospital stays and num
-
ber of discharges during the 26-year period prior to the study for par
-
ticipants and nonparticipants. The two groups were also compared for
hospitalization in 1998 as a marker of recent decline in health.
Siblings were identified by asking participants who gave DNA if
they had any brothers or sisters older than 88. The siblings were then
contacted in the same manner as described above and went through
the same questionnaire, tests, and collection of DNA.
36 JOURNAL OF AGING AND HEALTH / February 2001
Results
A total of 3,738 people were alive at the beginning of the recruit-
ment period, but 138 died before being contacted, leaving a total of
3,600 potential participants (see Figure 1). Of these, 2,262 (62.8%) con-
sented to participate in the study—1,814 (80.2%) by person and 448
(19.8%) by a proxy respondent. The reasons for using a proxy respon-
dent were dementia (57.0%), severe sensoric deficits (14.1%), unwill-
ingness to participate personally (14.1%), and illness (13.9%). In most
of the proxy interviews (77.9%), a first-degree relative was interviewed.
Most (86%) of the proxies saw the nonagenarian at least once a week.
Face-to-face interviews were conducted with the nonagenarian
alone in 81% of the cases, and in 19% of the cases, another person par
-
ticipated together with the nonagenarian, most often a child (49.9%)
or a spouse (16.8%). The interviewers considered the face-to-face
interview as easy to conduct in 64% of the cases, somewhat difficult in
26% of the cases, and difficult in 10% of the cases. Problems with
hearing (33%), vision (24%), and cognitive impairment (10%) were
the most common causes that impeded the interview.
There were 1,338 (37.2%) nonparticipants. Males were more likely
to participate than females (68.9% vs. 61.0%;
2
test, p < 0.001) (see
Table 1). There was no difference in residence type or marital status
Nybo et al. / THE DANISH 1905 COHORT 37
Figure 1. The ascertainment of participants from the Danish 1905 cohort survey.
Table 1
Demographic Characteristics at Baseline and Postsurvey Mortality of the 1905 Cohort in Denmark
Participants Participants All
by Person by Proxy Participants Nonparticipants Total
( n = 1,814, 50.4%)(n= 448, 12.4%) (n = 2,262, 62.8%) (n = 1,338, 37.2%) (N = 3,600, 100%)
Male Female Male Female Male Female Male Female Male Female
Number of participants 494 1,320 90 358 584 1,678 265 1,073 849 2,751
(% of sample) (27.2) (72.8) (20.1) (79.9) (25.8) (74.2) (19.8) (80.2) (23.6) (76.4)
Sex ratio (female:male) 2.67 3.98 2.87 4.05 3.24
Type of residence
House/apartment 300 757 21 77 321 834 142 547 463 1,381
(60.7) (57.3) (23.3) (21.5) (55.0) (49.7) (53.6) (51.0) (54.5) (50.2)
Sheltered housing/ 180 550 64 279 244 829 115 503 359 1,332
nursing home (36.4) (41.7) (71.1) (77.9) (41.8) (49.4) (43.4) (46.9) (42.3) (48.4)
Other 14 13 5 2 19 15 8 23 27 38
(2.8) (1.0) (5.6) (0.6) (3.3) (0.9) (3.0) (2.1) (3.2) (1.4)
Area of residence
Rural (less than 500 351 855 67 255 418 1,110 176 557 594 1,667
participants per square km) (71.1) (64.8) (74.4) (71.2) (71.6) (66.2) (66.4) (51.9) (70.0) (60.6)
Urban (more than 500 143 465 23 103 166 568 89 516 255 1,084
participants per square km) (28.9) (35.2) (25.6) (28.8) (28.4) (33.8) (33.6) (48.1) (30.0) (39.4)
Marital status
Widow/widower 318 1,068 54 297 372 1,365 140 853 512 2,218
(64.4) (80.9) (60.0) (83.0) (63.7) (81.3) (52.8) (79.5) (60.3) (80.6)
38
Divorced 12 51 — 16 12 67 7 54 19 121
(2.4) (3.9) (4.5) (2.1) (4.0) (2.6) (5.0) (2.2) (4.4)
Married 140 44 31 12 171 56 99 29 270 85
(28.3) (3.3) (34.4) (3.1) (29.3) (3.3) (37.4) (2.7) (31.8) (3.1)
Single 24 157 5 33 29 190 19 137 48 327
(4.9) (11.9) (5.6) (9.2) (5.0) (11.3) (7.2) (12.8) (5.7) (11.9)
Postsurvey mortality
a
31 55 20 55 51 110 38 120 89 230
(6.3) (4.2) (22.2) (15.4) (8.7) (6.6) (14.3) (11.2) (10.5) (8.4)
Note. Percentages in parentheses.
a. Within 6 months after the start of the survey.
39
among participants and nonparticipants (
2
test, p = 0.19, 0.13). As
expected, participants interviewed by proxy were more likely to live in
nursing homes. In all groups, females were more often nursing home
residents and bereaved of their spouses. Participants living in urban
areas (more than 500 persons per square km) were more likely to be
nonresponders (
2
test, p < .001), especially in the metropolitan area of
Copenhagen, where only 49% participated. Male participants were
more frequently widowers than were male nonparticipants (
2
test, p =
0.003).
The total death rate within 6 months was 8.9%. Mortality among
nonparticipants (11.8%) was higher than among participants (7.2%)
(log-rank test, p < .001). Mortality was especially high among persons
who refused to participate because of illness (18.6%) or hospitaliza
-
tion (44.4%). In all groups, the mortality among women was lower
than among men.
Information from the hospital discharge registry of Funen County
is shown in Table 2. Participants and nonparticipants were highly
comparable with regard to the number of hospital admissions (Mann-
Whitney test, p = .185), total number of days in hospital in the previ
-
ous 26 years (Mann-Whitney test, p = .58), and admissions during
1998 (
2
test, p = 0.571).
In all, 350 siblings to members of the 1905 cohort were eligible for
inclusion. The nonagenarians allowed us to contact 283 of them, and
we were able to trace 227 persons; 162 siblings (71.4%) participated.
A total of 1,639 (72.5%) participants from the 1905 cohort gave a
sample of DNA: 1,317 (80.4%) by means of blood spots and 322
40 JOURNAL OF AGING AND HEALTH / February 2001
Table 2
Hospital Discharge Registry Information on
Eligible Participants From the 1905 Cohort Living in the County of Funen
Participants Nonparticipants Total
Hospital Discharges (n = 283) (n = 106)(N= 389)
Number of hospital discharges 4 (0-19) 3 (0-17) 3 (0-19)
from 1973 to 1998, median (range)
Total number of days in hospital
from 1973 to 1998, median (range) 38 (0-299) 33 (0-265) 37 (0-299)
Number of persons hospitalized
in 1998 (%) 85 (30.0) 35 (33.3) 120(30.8)
(19.6%) as cheek swabs. Most of the blood spots—1,201 (91.8%)—
were classified as being of good quality. Every third cheek swab was
tested with a Polymerase Chain Reaction, revealing a fragment of 250
base pairs on all, indicating good quality of the DNA. In all, 137
(84.5% of the interviewed) siblings gave a DNA sample.
Discussion
The survey of the 1905 cohort has shown that it is possible to con
-
duct a large-scale genetic-epidemiological survey among extremely
old people in Denmark. To the best of our knowledge, a survey of a
national cohort of nonagenarians, including physical and cognitive tests
and collection of DNA material, has never been conducted before.
Comparisons of cross-sectional studies of octogenarians and cente-
narians indicate dramatic increases in the frequency of nursing home
admissions and in the occurrence of impairments, disability, and mor-
bidity across these age groups. However, nonagenarians represent an
age group whose cognitive, physiological, and physical features have
yet to be investigated systematically. We find nonagenarians very
important to study for several reasons. A large degree of heterogeneity
with respect to both physical and cognitive functioning can be ex-
pected among nonagenarians, and levels of functioning will often tend
to decrease rapidly as the nonagenarians age. Compared to centenari-
ans, nonagenarians are more functionally active and therefore more
able to participate in a survey and to carry out the test battery. There
-
fore, following nonagenarians over time may provide more valuable
information than following centenarians. Furthermore, what can be
learned from nonagenarians may be more generalizable to the total
older population than what can be learned from the very rare and
selected group of centenarians. We believe that the follow-up studies
of this age group within a few years can be expected to shed some light
on the aging process among the extremely old.
Studying the oldest old is a very complex issue; a major concern is
to maximize the participation rate (Simonsick et al., 1997). We took
several initiatives that have been reported to improve response rates
(Lundberg & Thorslund, 1996; Rodgers & Herzog, 1992). The letter
inviting the nonagenarians to participate in the survey was carefully
Nybo et al. / THE DANISH 1905 COHORT 41
composed and written in a large font. It was stressed that the survey
was home based and that participants could participate in the survey
without giving a sample of DNA. Furthermore, we encouraged prox
-
ies to participate together with or instead of the participant. The inter
-
viewers were well trained and motivated and made great efforts to
include the potential participants in the survey. Additional informa
-
tion letters were sent to the caregivers of the nonagenarians. Despite
these initiatives, the participation rate was not very high (62.8%). Sev
-
eral factors may have contributed to the lower participation rate in our
study compared to other surveys. First of all, identification of eligible
participants in previous studies may have been incomplete due to lack
of registration of the population and to age-validation problems
(Hoinville, 1983; Rodgers & Herzog, 1992). In Denmark, we have a
complete and valid registration of the entire population. Second, in
contrast to many population-based surveys where people with severe
cognitive or sensory impairments are excluded, no exclusion criteria
were used in the 1905 cohort survey to enhance representativeness.
Population-based studies of people age 90 and older with good par-
ticipation rates have been reported in the literature (Forsell et al.,
1995; O’Connor et al., 1989; Steen & Djurfeldt, 1993). However, they
are characterized by having been conducted in relatively small areas
with a limited number of participants and by using physicians or
nurses for the examination. This probably makes potential partici-
pants more committed to the survey and therefore less likely to be
nonresponders. Larger surveys of elderly have response rates similar
to ours (Ebly, Parhad, Hogan, & Fung, 1994).
Therefore, considering the valid identification of the entire 1905
cohort, the lack of exclusion criteria, the size of the study, and the age,
morbidity, and disability level of these extremely old people, we find
the relatively low participation acceptable.
In the 1905 cohort survey, we evaluated differences between partic
-
ipants and nonparticipants with valid register-based data. We found no
differences in marital status or housing type, but men and participants
living in rural areas were more likely to participate, which are findings
common to other studies of elderly people (Christensen et al., 1999;
Ganguli, Mendelson, Lytle, & Dodge, 1998; Sørensen, Sivertsen,
Schroll, & Gjorup, 1982). Analyzing hospitalization patterns did not
indicate that responders were more healthy than nonresponders, as
42 JOURNAL OF AGING AND HEALTH / February 2001
hospitalization patterns in the 26 years preceding the survey and in
1998 showed no difference between the two groups. However, in a
6-month period after the start of the survey, nonresponders had higher
mortality, indicating that terminal illness was one of the reasons for
nonparticipation. The higher nonresponse rate in the urban areas was
not caused by a higher mortality among persons living there compared
with persons living in rural areas; the mortality rate in urban areas was
10.8% versus 15.4% in rural areas. Furthermore, the Danish popula
-
tion of elderly is very homogenous with respect to ethnicity and social
conditions. We think these circumstances provide support for regard
-
ing the survey as reasonably representative of the population of the
oldest old. In certain parts of the country, the participation rate was
high (up to 73%), and it is possible to stratify on residency when ana
-
lyzing the data and thereby have an additional test of the representa-
tiveness of the overall sample.
A substantial proportion of the oldest old is incapable of providing
accurate responses to survey questions or even participating in the
interview due to cognitive impairment or frail health (Rodgers et al.,
1992). This necessitates the frequent use of proxy responders, as seen
in the present study. The literature on proxy information in epidemio-
logical surveys suggests that validity varies considerably depending
on the type of information sought (Rodgers et al., 1992) but can be
improved by using different kinds of proxies (Lundberg & Thorslund,
1996). In our study, a substantial proportion of the participants
(19.8%) were interviewed by proxy. The proxies knew the participant
very well (86% saw the participant at least once a week), and further
-
more, the interviewers were instructed to contact caregivers if there
were any questions concerning medication or ADL that the proxy was
unable to answer.
When choosing instruments for the survey, we had to take into con
-
sideration that the survey was home based and that the interviewers
had no medical background. Furthermore, the length of the visit
should be limited to no more than 2 hours, because very old people get
tired easily, which probably could impair the quality of information
retrieved.
Physical performance measures were included in the study to pro
-
vide objective and detailed information about functional capacity to
understand the pathway from pathology to disability. However, the
Nybo et al. / THE DANISH 1905 COHORT 43
expected high prevalence of disabilities and sensory deficits makes
many of the instrument batteries developed for studies of younger
elderly unsuitable for use in a cohort of very old people (Olsen,
Jeune, & Andersen Ranberg, 1996; Ravaglia et al., 1997). The physi
-
cal performance tests were therefore carefully selected in consider
-
ation of the age and expected functional capacity of the study popula
-
tion. The majority of the participants could perform the tests, and no
accidents occurred.
Twin studies indicate that the influence of genes increases with age
for a number of important attributes (McClearn et al., 1997). Compar
-
ison of DNA from the oldest old with DNA from younger individuals
is among the design options for identifying genes that are of impor
-
tance for the aging process and life span. To make future genetic stud
-
ies feasible, DNA from the nonagenarians and their siblings was col-
lected. The majority of the participants were willing to give a sample,
and the DNA was of good quality. The relatively low participation rate
in the survey will only represent a bias problem in a prevalence study if
the genetic factors studied are associated with both nonparticipation
and the outcome of interest but not for folllow-up studies within the
response group. The “affected sib pair design” (“affected” for longev-
ity), which we intend to use, is robust against selection problems
because this design intends to identify families with more than one
case (here, two siblings) (Khoury, Beaty, & Cohen, 1993).
In conclusion, despite the known difficulties of conducting surveys
among the extremely old, the Danish 1905 cohort survey shows that it
is feasible to conduct a nationwide genetic-epidemiological survey,
including collection of DNA, among nonagenarians. The large and
fairly nonselected sample of interviewed persons provides a unique
research resource for describing the population of nonagenarians and
for studying environmental and genetic determinants of loss of abili
-
ties and mortality, thereby helping to elucidate the aging process.
REFERENCES
Avlund, K., & Schultz Larsen, K. (1991). What do 70-year-old men and women actually do?
And what are they able to do? From the Glostrup survey in 1984. Aging Milano, 3, 39-49.
44 JOURNAL OF AGING AND HEALTH / February 2001
Christensen, K., Gaist, D., Jeune, B. & Vaupel, J. W. (1998). A tooth per child? [letter]. Lancet,
352, 204.
Christensen, K., Holm, N. V., McGue, M., Corder, L., & Vaupel, J. W. (1999). A Danish popula
-
tion-based twin study on general health in the elderly. Journal of Aging and Health, 11,
49-64.
Ebly, E. M., Parhad, I. M., Hogan, D. B., & Fung, T. S. (1994). Prevalence and types of dementia
in the very old: Results from the Canadian Study of Health and Aging. Neurology, 44,
1593-1600.
Folstein, M. F., Folstein, S. E., & McHugh, P. R. (1975). Mini-Mental State: A practical method
for grading the cognitive state of patients for the clinician. Journal of Psychiatric Research,
12, 189-198.
Forsell, Y., Jorm, A. F., von Strauss, E., & Winblad, B. (1995). Prevalence and correlates of
depression in a population of nonagenarians. British Journal of Psychiatry, 167, 61-64.
Ganguli, M., Mendelsohn, A., Lytle, M., & Dodge, H. (1998). A follow-up comparison of study
participants and refusers within a rural elderly population. Journal of Gerontology, 53,
M465-M470.
Guralnik J. M., Fried, L. P., Simonsick, E. M., Kasper, J. D., & Lafferty, M. E. (1995). The
women’s health and aging study: Health and social characteristics of older women with dis
-
ability (NIH Publication No. 95-4009). Bethesda, MD: National Institute of Aging.
Harris, J. H., Finucane, P. M., Healy, D. C., & Bakarich, A. C. (1997). Use of inpatient hospital
services by people aged 90-99 years. Medical Journal of Australia, 167(8), 417-420.
Heeren, T. J., Lagaay, A. M., von Beek, W. C., Rooymans, H. G., & Hijmans, W. (1990). Refer-
ence values for the Mini-Mental State Examination (MMSE) in octo- and nonagenarians.
Journal of the American Geriatric Society, 38, 1093-1096.
Hoinville, G. (1983). Carrying out surveys among the elderly. Journal of Market Research Soci-
ety, 25, 223-237.
Katz, S., Downs, T. D., Cash, H. R., & Grotz, R. C. (1970). Progress in development of the index
of ADL. The Gerontologist, 10, 20-30.
Khoury, M. J., Beaty, T. H., & Cohen, B. H. (1993). Fundamentals of genetic epidemiology.New
York: Oxford University Press.
Lundberg, O., & Thorslund, M. (1996). Fieldwork and measurement considerations in surveys
of the oldest old. Social Indicators Research, 37, 165-187.
McClearn, G. E., Johansson, B., Berg, S., Pedersen, N. L., Ahern, F., Petrill, S. A., & Plomin, R.
(1997). Substantial genetic influence on cognitive abilities in twins 80 or more years old. Sci
-
ence, 276, 1560-1563.
McGue, M., & Christensen, K. (1997). Genetic and environmental contributions to depression
symptomatology: Evidence from Danish twins 75 years of age and older. Journal of Abnor
-
mal Psychology, 106, 439-448.
Meller, I., Fichter, M., Schroppel, H., & Beck Eichinger, M. (1993). Mental and somatic health
and need for care in octo- and nonagenerians: An epidemiological community study.
European Archives of Psychiatry and Neurobiological Sciences, 242, 286-292.
Nagi, S. Z. (1976). An epidemiology of disability among adults in the United States. Milbank
Memorial Fund Quarterly of Health and Society, 54, 439-467.
O’Connor, D. W., Pollitt, P. A., Brook, C. P., & Reiss, B. B. (1989). A community survey of men
-
tal and physical infirmity in nonagenarians. Age Ageing, 18, 411-414.
Olsen, H., Jeune, B., & Andersen Ranberg, K. (1996). Hundredarige pa Fyn. Sygdomsforekomst
og funktionsevne. [Centenarians in the county of Funen. Morbidity and functional capacity].
Ugeskr Laeger, 158, 7397-7401.
Nybo et al. / THE DANISH 1905 COHORT 45
Ravaglia, G., Forti, P., Maioli, F., Boschi, F., Cicognani, A., Bernardi, M., Pratelli, L.,
Pizzoferrato, A., Porcu, S., & Gasbarrini, G. (1997). Determinants of functional status in healthy
Italian nonagenarians and centenarians: A comprehensive functional assessment by the
instruments of geriatric practice. Journal of the American Geriatric Society, 45, 1196-1202.
Ritchie, K., & Fuhrer, R. (1992). A comparative study of the performance of screening tests for
senile dementia using receiver operating characteristics analysis. Journal of Clinical Epide
-
miology, 45, 627-637.
Rodgers, W. L., & Herzog, A. R. (1992). Collecting data about the oldest old: Problems and pro
-
cedures. In R. M. Suzman, D. P. Willis, & K. G. Manton (Eds.), The oldest old (pp. 135-156).
New York: Oxford University Press.
Roth, M., Tym, E., Mountjoy, C. Q., Huppert, F. A., Hendrie, H., Verma, S., & Goddard, R.
(1986). CAMDEX: A standardised instrument for the diagnosis of mental disorder in the
elderly with special reference to the early detection of dementia. British Journal of Psychia
-
try, 149, 698-709.
Simonsick, E. M., Maffeo, C. E., Rogers, S. K., Skinner, E. A., Davis, D., Guralnik, J. M., &
Fried, L. P. (1997). Methodology and feasibility of a home-based examination in disabled
older women: The Women’s Health and Aging Study. Journal of Gerontology, 52, M264-
M274.
Sørensen, K. H., Sivertsen, J., Schroll, M., & Gjorup, S. (1982). A socio-medical population
study of the elderly in the city of Copenhagen 1978/79: General presentation and method-
ological aspects. Danish Medical Bulletin, 29, 274-280.
Steen, B., & Djurfeldt, H. (1993). The gerontological and geriatric population studies in
Gothenburg, Sweden. Zietschrift Für Gerontologie, 26, 163-169.
Vaupel, J. W., Carey, J. R., Christensen, K., Johnson, T. E., Yashin, A. I., Holm, N. V.,
Iachine, I. A., Kannisto, V., Khazaeli, A. A., Liedo, P., Longo, V. D., Zeng, Y., Manton, K. G.,
& Curtsinger, J. W. (1998). Biodemographic trajectories of longevity. Science, 280, 855-
860.
46 JOURNAL OF AGING AND HEALTH / February 2001
