| HOME | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
| ||||||||||||||||||||||||||||||||
RESEARCH ARTICLE |
a Gerontology Research Centre, Lund, Sweden
b Gerontology Center, University of Georgia, Athens
c Tokyo Metropolitan Institute of Gerontology, Japan
Bo Hagberg, The Gerontology Research Centre, Karl XII gatan 1, S-222 20 Lund, Sweden E-mail: bo.hagberg{at}psychology.lu.se.
Decision Editor: Toni C. Antonucci, PhD
 |
Abstract |
|---|
 TOP Abstract MethodsResultsDiscussionReferences |
|---|
MAINTENANCE of cognition has been found to be a crucial determinant for quality of life and successful aging (
Baltes and Baltes 1990
;
Palmore, Busse, and Maddox 1985;
Thomae 1976) as well as longevity and survival in old age (
Poon et al. 2000). In multivariate analyses, cognition has been related to health and survival in the Bonn Study (
Thomae 1976), the Duke Study (
Palmore 1985), the Seattle Study (
Schaie 1983), the Berlin Aging Study (BASE) (
Maier and Smith 1999), the Seattle Longitudinal Study (
Bosworth and Schaie 1999;
Bosworth, Schaie, and Willis 1999), and the Swedish Population Study (
Steen, Berg, and Steen 1998). How health and cognition are related in oldest old persons is, however, not quite clear. The relationship can be bidirectional. That is, either illness causes the cognitive decline, or the cognitive decline predicts disease. A causal structural model on survival of the Swedish centenarians seems to indicate that declining health affects cognition, which in turn predicts survival after 100 years of age (
Samuelsson et al. 1997).
Cognitive breakdown, as in the dementia illnesses, can be found as early as the 40s (
Brun, et al. 1990). It is interesting to note that two seemingly contradictory observations about cognition among very old persons have been made. On one hand, cognitive dysfunction was found to be associated with increased mortality and selective survival, resulting in decreased prevalence of cognitive disability and lower variability in cognitive performances among oldest old persons (Perls, Morris, Wee Lock Ooi, & Lipsitz, 1993). On the other hand,
Schaie and Willis 1991 reported that inter- and intravariability in cognition seem to increase with advancing age. One of our aims in this article is to examine variability in cognitive performances in oldest old persons. Three possible outcomes exist. First, selective survival could result in lower variabilities in oldest old persons. Second, differences in normal age decline and attrition due to dementia and other central or peripheral diseases (e.g.,
Baltes and Lindenberger 1997;
Lindenberger and Baltes 1994) could produce higher variability in oldest old persons. Third, higher variabilities could be expected in the more robust semantic or experience-related functions, whereas lower variabilities could be explained by the more rapid decline to floor effects in the process-related functions during the aging process. This phenomenon would be most clearly demonstrated in very old persons, that is, centenarians (
Christensen et al. 1994;
Salthouse 1989).
Related to the description of cognition is the problem of how to assess and estimate the number of dementia cases among very old persons. The prevalence rates of dementia have been shown to increase dramatically with age. Five percent of the population aged more than 65 years are afflicted, and the percentage increases to 20 for the population aged more than 80. The prevalence rates for each successive age group after 60 are estimated to double every 5 years (
Jorm, Korten, and Henderson 1987). A summary of eight European studies (
Hofman et al. 1991) showed a dramatic increase of dementia prevalence rate from 40% to 70% between the ages of 90 and 95. If these estimates are correct, then the prospect of an individual's reaching 100 years without a dementia diagnosis seems unlikely. However, results from longitudinal studies show an increased mortality rate among dementia patients. The mortality rate among old persons with organic mental impairment was 46% after 4 years and 81% after 8 years with the disease, compared to 17% and 38% among nonorganic patients (
Cooper and Bickel 1989). These figures support the idea of "the survival of the healthy." The prevalence figures for dementia among the centenarians could be expected to deviate from the trajectories calculated from younger people (
Reischies et al. 1997). In other words, attrition from dementia because of premature death would reduce the prevalence of dementia among the centenarians in the population to a much lower figure.
Ravaglia and colleagues 1999 reported in an Italian study a prevalence of 62% for 92-year-old persons, and
von Strauss, Viitanen, De Ronchi, Winblad, Fratiglioni 1999 in a recent Swedish study gave a prevalence figure of 48% for persons aged 95 years and older. von Strauss and colleagues also noted that the prevalence rate is higher for women. This information is thought to reflect the differential distribution of dementia risk, which raises a related question. How should the conventional diagnostic criteria for dementias, as defined by the revised third edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM-III-R;
American Psychiatric Association 1987) or the International Classification of Diseases (ICD-10), be applied to the centenarians?
Wettstein 1999, for instance, brought up the question of how much allowance should be given to natural aging decline before considering a diagnosis of dementia. And, what prevalence figures are valid for the centenarian group?
In this paper we aim to provide a better understanding of centenarians' cognition by comparing and contrasting findings from the Swedish, U.S., and Japanese centenarian studies in three areas: (a) definitions and prevalence of dementia between Japan and Sweden, (b) levels of cognitive performances between centenarians and younger age groups, and (c) interindividual variations in cognitive performances of centenarians and younger persons in Sweden and the United States.
|
Methods |
|---|
|
TOPAbstractMethodsResultsDiscussionReferences |
|---|
Japan
Sampling and recruitment
In the study period of 1987–89, 218 of the 509 centenarians living in the Tokyo metropolitan area at that time (
Fig. 1) were interviewed by a team consisting of one psychiatrist and one psychologist, who used a semistructured interview form on a door-to-door basis (
Homma, Nakazato, and Shimonaka 1990;
Homma, Shimonaka, and Nakazato 1992). Study planners intended to interview all the centenarians in Tokyo; consent for the interviews was obtained from approximately 43% (n = 218). Thus, it should be kept in mind that the participants were not representative in terms of random sampling. Because dementia prevalence studies of centenarians are rare, we included this study in the comparison.
|
) was used by a psychiatrist to assess the severity of dementia. The CDR is derived from a semistructured interview with the patient and an appropriate informant. The CDR rates impairment in each of six cognitive categories (memory, orientation, judgment and problem solving, community affairs, home and hobbies, and personal care) on a 5-point scale, where none = 0, questionable = 0.5, mild = 1, moderate = 2, and severe = 3. From the six category ratings, the global CDR is established by indexing scores where a CDR of 0 equals no dementia, and a CDR of 0.5, 1, 2, or 3 indicates questionable, mild, moderate, or severe dementia, respectively.
Dementia diagnosis
The CDR has high interrater reliability for physicians (
Burke et al. 1988) and nonphysicians (
Forsell, Fratiglioni, Grut, Viitanen, and Winblad 1992;
McCulla et al. 1989) and the six categories for staging are directly linked to validated diagnostic criteria following the DSM-III classification (
Morris, McKeel, Fulling, Torack, and Berg 1988). Along with the CDR, the Hasegawa Dementia Scale (HDS;
Hasegawa 1983;
Hasegawa, Inoue, and Moriya 1974), a brief psychometric test, was administered by a psychologist. Hasegawa and colleagues developed the HDS to screen and to judge the severity of dementia. The score ranges from 0, indicating complete impairment, to 32.5, the maximum score. A score of 0–10 points indicates severe cognitive impairment, corresponding to severe dementia. The HDS has been one of the most widely used psychometric tests in Japan (
Homma and Hasegawa 1989;
Homma et al. 1990).
Comparison groups
In an epidemiological survey conducted in the Tokyo metropolitan area in 1980, the population of people aged 65 years and older was 1,101,144. Of them, researchers selected 4,568 persons randomly and screened them to find persons suffering from dementia in different age groups. One hundred and fifty-five 70-year-old persons were derived from the sample.
Sweden
Sampling and recruitment
Between 1987 and 1992, a Swedish centenarian study, which examined a local population drawn from the National Register, was in progress. Use of the register secured two things: first, that the age of the individual was correct, and second, that all individuals in the cohort were located. The register also showed the rapid increase in the number of centenarians, which has doubled during the last 8 years and is now 38 per million inhabitants. Participants were five cohorts of centenarians living in southern Sweden and born between 1887 and 1892. Each centenarian in this area was invited to the investigation by letter and examined within 6 months after his or her 100th birthday. In addition, the closest relative was also invited to take part.
A total of 164 individuals were available for the study ( Fig. 2). Twenty-one were lost during this period (dead or lost records for unknown reasons), leaving 143 individuals. Seventy percent of those individuals were examined (n = 100); 82 were women and 18 were men. Data were collected through home visits or, in cases of institutional care, at the institution. A close relative was in most cases present and also interviewed (n = 86). The interviewing team consisted of a physician, a psychologist, and a nurse. The first visit included medical and sociological assessments plus an interview with the relative. The second visit included psychological assessments of the centenarian. Seventy-one percent had 7 years of education, and 29% had 8 or more years of education; 62% were living in the community, in either their own or relatives' homes, or old-age homes.
|
), Digit Span Forward and Backward (
Wechsler 1981), Memory for Designs (
Graham and Kendall 1960), and the Five Objects test, as used in clinical practice. In the Five Objects test, the participant must remember the previously shown common objects, both with and without distraction. In addition, a learning retention test was used, where learning was defined as the number of explanations a participant needed to comprehend the meaning of previously unknown words and retention was defined as the number of words he or she recalled at the end of the entire test session. Simple reaction time, clinical assessment (
Hagberg and Gustafson 1985), interviews about quality of life (
Nordbeck, Alfredson, Hagberg, Samuelsson, and Samuelsson 1991), and personality ratings were also recorded.
Dementia diagnosis
A functional dementia screening was made in terms of autonomy-dependency with the
Berger 1980 scale. This scale measures degree of autonomy in five steps on the basis of how much instruction and assistance is needed for daily life functions. The diagnosis of dementia was completed by psychological tests according to DSM-III-R criteria with the following instruments for assessing Categories A through E. Participants with dementia showed
); further signs of aphasia and agnosia were assessed by Eisenson's aphasia test (modified version according to
Hagberg and Ingvar 1976). ) together with >5 behavior disturbances in test performance.
Comparison groups
Four comparison groups were chosen. One was from an ongoing longitudinal study of a representative sample of men born in 1914, examined at the age of 68 years (
Steen, Hagberg, Johnsson, and Steen 1987) and again at the age of 82, with a comprehensive interdisciplinary assessment battery (André-Petersson, Hagberg, Hedblad, Janzon, & Steen, 1999). Two other comparison groups were from the Swedish Longitudinal Study, H-70; data from two measurement points, at ages 70 and 75, were used (
Berg 1980). A fourth group was included from the population study Lund 80+ (
Svensson, Dehlin, Hagberg, and Samuelsson 1993), where comparisons were made with data from the age of 80.
Georgia, USA
Sampling and recruitment of centenarian and comparison groups
About 100 participants in each of three panels of cohorts participated in the study, which began in 1988. The panels were 60–69 years (M = 64.9), 80–89 years (M = 82.6), and 100 years or older (M = 100.7). Participants were recruited from the southeastern United States, primarily from the state of Georgia. All subjects were community dwelling, cognitively intact, and in reasonably good health. A community-dwelling person was defined as anyone who was self-sufficient, or partially self-sufficient, living in the community and not in a custodial institution. Persons who lived in their relatives' homes, group or old-age homes, and life care communities were considered community-dwelling individuals in this study (Poon, Martin, Clayton, Messner, & Noble, 1992).
A cognitively intact person was defined as one not demented or disoriented. Inclusion criteria were Mini-Mental State Examination (MMSE;
Folstein, Folstein, and McHugh 1975) score greater than 21, and Global Deterioration Scale (GDS;
Reisberg et al. 1984) not lower than Stage 2. Sensory or physical handicaps or psychiatric problems were noted, but these participants were not excluded from the study. Thirty-three percent of the participants were men, 66% were women, 73% were Caucasian Americans, and 27% were African Americans. These are representative proportions for gender and race in the state of Georgia. Thirty-four percent of all centenarians had less than 9 years of education, compared with 21% in the octogenarian and 15% in the sexagenarian groups.
To obtain a representative sample, the study researchers enlisted the services of the University of Georgia Survey Research Center for the recruitment of the sexagenarians and octogenarians. Centenarians were recruited by referrals from a wide network of state and local agencies, churches, television, and print media. Individuals were also contacted directly by the project staff. All centenarians were tested in their homes. Octogenarians and sexagenarians were tested in groups, usually at a community center or a college near their homes.
Cognitive assessment
Detailed information regarding participants' characteristics, recruitment and testing procedures, and instruments are described in
Poon and colleagues 1992. Eleven primary substantive domains were evaluated. Among these, mental health and cognition were assessed with subtests from WAIS: Vocabulary, Block Design, Arithmetic, and Picture Arrangement. In addition, Paired Associate Learning, President Recall, and Recognition were used. Hypotheses regarding direct and indirect effects of these domains to successful adaptation of the 60-, 80-, and 100-year-old cohorts are described in
Poon and colleagues 1992.
Statistical Analyses
To compare variability in measurements of cognition used in the different studies in a way that particularly highlights the shape of the distribution (
Pagano 1994;
Weller and Romney 1990), we constructed a scale and level independent quotient: (Q3-Q1)/Range, where Q1 and Q3 were first and third quartiles and the range was the maximum–minimum difference. The quotient was thus a measurement of variability relative to the observed range. The possible values of the quotient were between 0 and 1, with higher values indicating distributions with greater spread; a value close to 0 indicated distribution centered around the medium, whereas a value close to 1 indicated distribution with many values at the extremes. Dementia prevalence was expressed in percentages of the total centenarian group and of each gender. Interaction between measures in the Japanese study was analyzed with variance analysis.
|
Results |
|---|
|
TOPAbstractMethodsResultsDiscussionReferences |
|---|
|
|
estimated an exponential model to provide age-specific prevalence rates of dementia. According to the model, prevalence rates of multi-infarct dementia doubled every 5.3 years and rates of dementia of the Alzheimer type doubled every 4.3 years. Although it was difficult to diagnose types of dementia in centenarians, the total prevalence rate of 63% in the Japanese centenarians seems to be compatible with the model. There was also a marked gender difference for the prevalence of dementia (
Fig. 1). Of 218 centenarians, 43% of the men were demented, whereas as many as 71% of the women were demented (Table 2 ). The difference was most remarkable in those with severe dementia; 3% of the men and 27% of the women were severely demented. The higher prevalence rate of dementia in women has also been found in the previous geropsychiatric epidemiological surveys on dementia (
Homma and Hasegawa 1989).
Comparison between psychometric assessment (HDS) and rated impairment (CDR)
Table 3 shows the distribution of the mean scores of HDS, a performance-based psychological test. The data from 172 centenarians were evaluated. The mean HDS score among those without dementia was 26.2 (SD 4.1), and the mean scores decreased with advancing dementia.
Karasawa, Kawashima, and Hasegawa 1975 noted that cognitive functioning in centenarians differed qualitatively from that of elderly persons aged 70–80 years and suggested that the cognitive decline of the centenarians is partly attributable to disuse of everyday functioning. Seeking to confirm these findings, we compared HDS scores of a sample of 70-year-old participants to the scores of the centenarians in the present study.
|
|
Swedish Analyses
Prevalence of dementia
Twenty-seven percent of the centenarians were classified as mildly, moderately, or severely demented according to the DSM-III-R criteria. Of the 82 women, 30% were demented, and of the 18 men, 16% were demented. Yet, including the drop outs (when examination was refused either by the centenarian or by a relative, or when reference was made to a diagnosis of dementia as an excuse for not taking part in the study), the prevalence of dementia for the whole group was estimated to be 40 ± 10% (
Fig. 2). This figure is lower than the figures of prevalence for dementia in other centenarian studies, as, for example, in the Hungarian study (
Beregi and Klinger 1989) showing 43% for men and 63% for women, and the Japanese study with 63% (see previous section).
The wide variation in prevalence figures for dementia in the centenarians may be explained by differences in the criteria used, differences in measurement variables, and/or differences in the sampling of participants to be studied. These various explanations for the differences in observations are considered further in the Discussion.
Cognitive functioning
Because of diminished hearing (20%) or reduced eyesight (21%), psychometric assessment was sometimes difficult, and consequently the number of observations was reduced to between 43 and 76 persons for various cognitive tests (
Samuelsson et al. 1997). In verbal performance, the centenarians scored significantly lower than the nonpatient groups aged 16–34, 36–57 and 56 years (
Hagberg and Ingvar 1976). The variation was quite wide, with some centenarians performing as well as the average younger person (see
Fig. 4). A statistically significant difference was also shown in the Digit Span Forward (= 4.2) and Backward (= 2.6) tests (p < .001) comparing centenarians with 70-, 75-, and 80-year-old individuals (
Berg 1980;
Svensson et al. 1993). Also in this test, the interindividual variation was wide, and a subgroup of centenarians had reached their old age with very little cognitive decline (see
Fig. 4). Comparing the results from these two tests also appeared to support the earlier findings that crystallized intelligence is better preserved in old age and that processing aspects and speed-related functions are more likely to be affected negatively (
Ansley, Stankov, and Lord 1993;
Smith and Baltes 1993;
Svensson, Dehlin, Hagberg, and Samuelsson 1993). On the basis of these results, lower performances in some older adults may occur for some groups of individuals, but lower performances are not inevitable. The lower test scores may relate to health status as much as to chronological age.
|
Variability in the U.S. Data
We examined the variability in different cognitive abilities in the U.S. data using a measure of interquartile differences in relation to range. (See the Statistical Analyses section.) This was a measurement of variability relative to the observed range, independent of performance level, which also could be used to compare results from tests with different scaling properties. This measurement made it possible to compare the score distribution, not only between different age groups, but also between different cognitive functions, such as knowledge-based and process-based cognitive functions.
Table 4 shows measures of variability between 60-, 80-, and 100-year-olds. Two dichotomous patterns emerged. For those semantic or experience-related abilities that have been shown to be quite stable over the adult life, that is, maintenance of ability is expected (e.g., vocabulary and tertiary memory measured by President Recall and Recognition), we observed an increase in variability with increasing age. Some of the oldest old persons maintained these abilities, whereas others showed a dramatic decline. In contrast, for those process-related abilities where lowering of performance with age is expected (e.g., block design, arithmetic, and paired associate learning), a decrease in variability was observed as frailty sets in with oldest old persons. A downward trajectory of performances was observed with increasing age, which contributed to decreased variability. Most likely, the decreased variability was associated with the approaching of the lower ability limit. Thus the data seem to show that the variability in cognitive abilities in oldest old persons was associated with the relative robustness of the function to the aging process, which in turn was associated with the qualitative differences in the cognitive functions.
|
).
|
|
|
Discussion |
|---|
|
TOPAbstractMethodsResultsDiscussionReferences |
|---|
Definition and Prevalence of Dementia
The prevalences of dementia reported in the Japanese and Swedish studies are quite different—63% in the Japanese study, with 43% men and 71% women, and 27–40% in the Swedish study, with 16% men and 30% women. The U.S. study excluded participants with dementia. In other centenarian studies, the reports of dementia prevalence also differ. In the Hungarian study, the prevalence of dementia is 43% for men and 63% for women (
Beregi and Klinger 1989). The Finnish study reported 20% for men and almost 50% for women (
Louhija 1994). There are at least two possible reasons for the discrepancy across studies: the definition of dementia and the sample representativeness employed. A primary reason for the different prevalence figures may be the use of different criteria in defining dementia in oldest old persons. Questionnaires with normative cutoff scores were used in the Japanese study. DSM-III-R criteria in terms of psychometric tests and clinical assessment were used in the Swedish study. In both cases, the paramount question is whether a compensation for aging decline was made in applying the normative cutoff. This was a practice in the Japanese study (
Fig. 3). Comparing the dementia staging according to the CDR with performance abilities on the HDS, we find that the same HDS score (e.g., 7) places a 70-year-old at Stage 3 but a centenarian at Stage 2. A similar bias can be shown with a moderate degree of dementia. Even when allowance is made for a more relaxed criteria for dementia for centenarians in the Japanese study, Japan's prevalence rate is still higher compared with that of Sweden and other countries. If this allowance is not made, the prevalence figures are even higher. Another possible explanation for high prevalence is the lack of representativeness. Because the Japanese study had a participation rate of 43%, it is more likely that there is selectivity in the sample. As the selection mechanisms are not known, it can only be speculated as to whether dementia might be overrepresented in the sample taking part in the study. Sometimes it is easier to recruit patients and participants living in institutions.
Inspection of the distribution for severity of dementia by gender, we find that the Japanese results show a much greater prevalence of severe dementia in women, 27.1% versus 3.2% for men (Table 2 ). A similar gender crossover for physical health was found by
Karasawa, Kawashima, and Kasahara 1979. Another similar phenomenon is found in racial crossover in mortality. The risk of death is much higher among younger African Americans than Caucasians, whereas it is lower among very old African Americans than among very old Caucasians (
Johnson Jackson and Perry 1989).
The lowest dementia prevalence rates reported so far come from the Swedish study. These figures are based on a geographical population of centenarians with a 70% participation rate. Among the examined centenarians, only 27% met the criteria for dementia. The observation and test performance are related to the DSM-III-R criteria. In a second analysis, the prevalence was evaluated in the total population, including the dropouts. This was possible because the reason was assessed for nonparticipation for all but 10 participants. For 5 participants, dementia was given as reason for nonparticipation. For 28 participants, refusals were due to disease or frailty. Assuming that all nonparticipants suffered from dementia, the prevalence goes up to 40%. This is clearly an overestimation, so it can be concluded that at most 40% of the population suffered from dementia.
The data in the present study highlight the difficulties of assessing dementia in this very old group. Whether a normative procedure or criteria-related procedure is used, the uncertainty of normative age-related decline in oldest old persons will result in different prevalence rates of dementia. The increase in interindividual variability in cognition also makes dementia assessment difficult. The Swedish data suggest that a reasonable overall estimation of dementia among centenarians is 40%.
Cognitive Performance Level
It is not surprising that all three centenarian studies report lower cognitive performances among centenarians compared with younger controls. The Swedish data show larger age differences in measures of fluid intelligence than in crystallized intelligence. In the Japanese study, cognitive performance was assessed with two scales, the HDS and the CDR. Comparing the outcome with younger age groups, we find that centenarians show not only lower performance level, but also a qualitative difference in cognitive performance and processing. Data obtained with the HDS show that centenarians, when compared with 70-year-olds, are less efficient in such abilities as spatial orientation, time estimation, immediate recall, counting backward, and subtracting by 7. These abilities are closely associated with fluid component of intelligence. The results are consonant with the general model of differential decline (
Cattell and Horn 1978;
Horn 1981;
Salthouse 1989;
Schaie and Willis 1991), with crystallized intelligence less vulnerable than fluid intelligence to decline with increasing age. In contrast to the use of questionnaires to assess cognitive level, the Swedish and U.S. studies used psychometric tests. The results from both the Swedish and the U.S. studies support the finding that larger age differences were found in process or fluid-related abilities compared with semantic or experientially related abilities. These two studies examined centenarians from a similar birth cohort, 1887–92, although the samples were different culturally.
Interindividual Variability
It is generally found that variability in cognitive performances increases with increasing age (
Schaie and Willis 1991). It has also been hypothesized that centenarians would follow a similar trend, with the largest cognitive performance variability. This hypothesis is partly supported by the Swedish and U.S. studies. That is, when a cognitive process tends to be stable over the adult life span, increased variability is found with oldest old persons because some centenarians were able to maintain these functions and others were not, owing to increased frailty, severe loss of peripheral senses, and diseases. When a cognitive process tends to show a downward trajectory over the adult life span, performance variability is found to decrease for centenarians as performance level reaches the lower limits. These findings of within-group variability are consistent with the differential decline of process or fluid and experiential or semantic abilities.
Given the range in mental performance both between and within individuals, the question was raised whether there are specific characteristics among the centenarians, either in the present general status or in their life histories, that could account for this diversity. Because centenarians are the longest living segment of the population, this question has to be dealt within the context of survival. Several studies have shown that cognition is one of the survival determinants (
Deeg, Hofman, and van Zonneveld 1990;
Palmore et al. 1985;
Samuelsson et al. 1997;
Thomae 1976).
Examination of extant longevity predictive models shows that many predictors depend on the maintenance of cognitive ability. For example,
Pitskhelauri 1982 showed that longevity predictors in centenarians are close family ties, personal hygiene, moderation in sleep and diet, physical activity, keeping up with daily routines, organized purposeful behavior, discipline and hard work, freedom and independence, balanced diet, love of family, low ambitions, and belief in God. Most of these demands have as prerequisite an intact cognition. It is interesting to note that when cognition declines, as in dementia, the death rate goes up and doubles in a 4-year period. The Swedish study (
Samuelsson et al. 1997) found that intactness of cognition was correlated with degree of independence, perceived good life span quality, higher systolic and diastolic blood pressure, better activities of daily living functioning, better hearing and sight, low number of hip fractures, high occupational status, better education, and longer living in cities. The U.S. study (
Poon et al. 2000) showed a strong relationship between cognition and mental and physical health, which are related to survival of oldest old persons as well as number of days of survival after an individual reaches the age of 100. Many of the activities referred to previously require good cognition. That these activities also promote the maintenance of cognitive performance is assumed, but not yet proven. The conclusions from the Swedish and the U.S. studies lead us to assume that there is an interaction between activity variables and cognition, which are of importance for both reaching and enhancing well-being at 100 years of age. The interaction also increases the prognosis for survival after 100 years. The finding that cognition has a positive correlation to blood pressure and survival is intriguing. Samuelsson and his colleagues (1997) noted that relatively high blood pressure (>152 mm/Hg systolic) is advantageous for the very old. In younger age groups, high blood pressure is considered a risk factor for cerebrovascular and cardiovascular diseases. In centenarians, this is hypothesized to have an opposite effect—a positive factor for mental fitness as well as for well-being and survival. Samuelsson and his colleagues further postulated that higher blood pressure enhances cognitive functioning, which in turn has a positive effect on survival. On the other hand, low blood pressure is found to be more prevalent in patients with Alzheimer's disease and a tentative risk factor for dementia in general (
Deeg et al. 1990;
Passant et al. 1996).
|
Acknowledgments |
|---|
Received for publication January 27, 1999. Accepted for publication July 6, 2000.
|
References |
|---|
|
TOPAbstractMethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  B. Hagberg and G. Samuelsson Survival After 100 Years of Age: A Multivariate Model of Exceptional Survival in Swedish Centenarians J. Gerontol. A Biol. Sci. Med. Sci., November 1, 2008; 63(11): 1219 - 1226. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. F. Calvert Jr., J. Hollander-Rodriguez, J. Kaye, and M. Leahy Dementia-free survival among centenarians: an evidence-based review. J. Gerontol. A Biol. Sci. Med. Sci., September 1, 2006; 61(9): 951 - 956. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Y. Gondo, N. Hirose, Y. Arai, H. Inagaki, Y. Masui, K. Yamamura, K.-i. Shimizu, M. Takayama, Y. Ebihara, S. Nakazawa, et al. Functional status of centenarians in Tokyo, Japan: developing better phenotypes of exceptional longevity. J. Gerontol. A Biol. Sci. Med. Sci., March 1, 2006; 61(3): 305 - 310. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. T Perls The different paths to 100 Am. J. Clinical Nutrition, February 1, 2006; 83(2): 484S - 487S. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. C. Antonucci Introduction to Special Section on Centenarians and Dementia J. Gerontol. B. Psychol. Sci. Soc. Sci., May 1, 2001; 56(3): P133 - P133. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||
| HOME | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
