| HOME | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
| ||||||||||||||||||||||||||||||||
RESEARCH ARTICLE |
1 Department of Psychology
2 Leonard Davis School of Gerontology, University of Southern California.
3 Department of Medical Epidemiology, Karolinska Institutet, Stockholm, Sweden.
4 Department of Psychology, University of Göteborg, Sweden.
Address correspondence to Margaret Gatz, PhD, Department of Psychology, University of Southern California, 3620 McClintock Avenue, Los Angeles, CA 90089-1061. E-mail: gatz{at}usc.edu
 |
Abstract |
|---|
 TOP Abstract MethodsResultsDiscussionREFERENCES |
|---|
The possibility that engaging in certain types of activities may protect against or delay onset of dementia has important implications regarding both prevention of dementia and mechanisms of the disease. The idea that mental activity is good for the mind in a similar manner to that in which physical activity is good for the body has been popularized with the phrase "use it or lose it." As Salthouse (1991)
pointed out, this idea is far from being new, with versions of the disuse perspective having been around since the very first studies of age and cognition in the 1920s. However, 80 years later, researchers are still trying to determine whether various lifestyles and certain activities may affect cognitive functioning in later life.
One of the most consistent findings in Alzheimer's disease risk factor research has been the association between low education and increased risk of the disease (see Cummings, Vinters, Cole, & Khachaturian, 1998, for a review of risk factors). Although the mechanism of this association is not known, the concept of cognitive reserve (Mortimer, 1997) has been used to explain these findings. Cognitive reserve is thought to affect the amount of brain deterioration a person can have before actually showing any signs of obvious impairment. There are three major views of how education may be related to cognitive reserve (Gatz et al., 2001). First, differences in education may reflect innate differences in cognitive reserve, such as genetically mediated differences in cognitive abilities. Second, education may be indicative of early life factors, such as socioeconomic status, which could affect both level of education and level of cognitive reserve (e.g., early nutrition and brain development).
Third, educational attainment may reflect mental stimulation during a critical period of brain development. The possibility that mental stimulation may increase synaptic density and, therefore, increase cognitive reserve has been raised (Katzman, 1993). Although still speculative, this use-it-or-lose-it perspective is supported by animal research, which has shown that increased stimulation leads to greater dendritic complexity (Lucassen, van Someren, & Swaab, 1998). These findings have led to the hypothesis that continued mental stimulation subsequent to the school years, either through occupational or leisure activities, may also lead to greater cognitive reserve and delay onset of dementia.
Relative to the number of education–dementia investigations, few studies have examined leisure activities and dementia. However, all of these studies have reported significant associations between activities and dementia. Unknown, of course, is whether other studies have failed to uncover an association and have fallen prey to publication bias. Five published studies have examined the relationship between activities and dementia in general or Alzheimer's disease (Fabrigoule et al., 1995; Friedland et al., 2001; Kondo, Niino, & Shido, 1994; Scarmeas, Levy, Tang, Manly, & Stern, 2001; Wilson et al., 2002). One study, by Friedland and colleagues (2001), categorized leisure activities for participants aged 20 to 60 into three domains: intellectual, passive, and physical activities. Friedland and colleagues found that greater activity in each of the three categories was associated with lower risk of Alzheimer's disease, even after controlling for education. Similarly, Scarmeas and colleagues (2001) found that intellectual, physical, and social activities were all related to a reduced risk of dementia.
In a Japanese sample, Kondo and associates (1994) found that a number of behaviors reflective of reduced leisure activities were associated with an increased risk of Alzheimer's disease. These researchers matched controls to cases by gender and age, but they did not control for education in their analyses. The risk factors found in this study were primarily social: not sending letters or calling, not visiting friends or relatives, rarely leaving home, being unsociable, and not having friends visit. One nonsocial item, rarely reading books or papers, was a risk factor for Alzheimer's disease (Kondo et al., 1994). Supporting the finding of a link between social factors and dementia, Fratiglioni, Wang, Ericcson, Maytan, and Winblad (2000) found that having limited social networks increased the risk of dementia in a Swedish sample. Both the Japanese study and the study by Friedland and colleagues obtained information from informants after the dementia had been diagnosed, a method that can be vulnerable to recall bias. These studies also collected information in a different manner for controls than for cases (Friedland et al., 2001, used self-reports for controls; Kondo et al., 1994, used a combination of self-report and informant report for the controls), which could have led to differences in the accuracy of information for controls compared with that for cases.
In a 3-year longitudinal study of all dementias combined, Fabrigoule and colleagues (1995) found that traveling, doing odd jobs or knitting, and gardening were found to reduce the risk of dementia. This study circumvented the problem of gathering retrospective information from informants by using self-report of current activities from participants who were not demented at baseline. However, although the authors controlled for age and baseline cognitive performance, they did not control for level of education. In addition, the follow-up time of 1 to 3 years makes the interpretation of this study complicated. Activity levels in those who became demented may have already decreased as a result of prediagnosis symptomatology (Katzman, 1995). That is, people who were diagnosed with dementia 1 to 3 years after baseline were probably experiencing subtle changes at baseline that may have already affected the types of activities in which they were involved. Subsequent longitudinal studies of leisure activities and dementia, with mean follow-up times of 2.9 years (Scarmeas et al., 2001) and 4.5 years (Wilson et al., 2002), also had this potential problem. Schooler and Mulatu (2001) found that leisure activities and intellectual functioning have reciprocal effects. Initially higher levels of intellectual functioning predicted greater activity, and greater activity led to higher levels of functioning. This finding underscores the difficulty in interpreting associations found between cognitive decline and leisure activities. Kondo and colleagues (1994) attempted to deal with this problem by inquiring about participants' behaviors 15–20 years prior to dementia onset, whereas Friedland and colleagues (2001) obtained reports about participants' activities before the age of 60. However, the information was entirely retrospective, and reports may have been influenced by intervening events.
The present study had the unique advantage of using data collected from the participants themselves more than 20 years prior to evaluation for dementia in order to examine whether participation in certain types of leisure activities during early adulthood was associated with subsequent risk of dementia. We predicted that greater participation in activities that were relatively more intellectual or social would reduce risk of dementia. We controlled for education as a means of adjusting for initially higher levels of intellectual functioning. To our knowledge, this study is the first to examine the relationship between leisure activities and dementia by use of a co-twin control design and represents the longest interval yet reported between gathering of leisure activity information and occurrence of dementia.
The co-twin control method is a matched pair analysis that takes twin pairs who are discordant for the disease to find whether the diseased twin was more often exposed to a given risk factor or less often exposed to a protective factor. The major advantage of this design is that the matching of pairs is better than for any other type of case-control study. Monozygotic (identical) twins share 100% of their genes, and dizygotic (fraternal) twins share 50% of their segregating genes; thus, this design controls for genetic and familial factors. In addition, given same-sex twin pairs, cases and controls are identical with respect to gender.
|
METHODS |
|---|
|
TOPAbstractMethodsResultsDiscussionREFERENCES |
|---|
) who both completed leisure activity items in 1967 and were assessed for dementia under the aegis of either the Study of Dementia in Swedish Twins (Gatz et al., 1997) or the OCTO-Twin study (McClearn et al., 1997). One concern with a study of twins is whether results can be generalized to the larger population. Providing evidence for the representativeness of the present sample, an earlier examination of the Swedish Twin Registry and a recent analysis of the OCTO-Twin data found that Swedish twins are similar to Swedish nontwins of the same age in terms of health status and sociological characteristics (Lichtenstein et al., 2002).
Swedish Twin Registry (STR)
All like-sexed twins born between 1886 and 1925 were sent a questionnaire between 1960 and 1961 if both twins were alive and living in Sweden at the time. Those who responded to the questionnaire were enrolled in the STR as the "old cohort" and were sent a questionnaire in 1967 that included items on different types of leisure activities (N = 20,770). The response rate for the activity items was 81%.
OCTO-Twin study
The OCTO-Twin study is composed of all possible pairs of twins from the STR who were at least 80 years old and both alive at the time that this study began in 1991 to 1994 and agreed to participate (N = 702). This represents a response rate of 86%. The twin pairs were subsequently followed at three additional waves conducted continuously over 2-year intervals. Dementia cases were identified at each wave. The present study includes all members of the OCTO-Twin study who completed the leisure activity items in 1967 (n = 640).
Study of Dementia in Swedish Twins
Beginning in 1987, the Study of Dementia in Swedish Twins identified dementia cases from a defined subset of the STR known as the Swedish Adoption/Twin Study of Aging (SATSA; Pedersen et al., 1991). SATSA consists of all twin pairs who reported having been reared apart and a matched sample of twins who were reared together. Between 1987 and 1991, SATSA members were screened and evaluated for dementia. In addition, because members of SATSA participated in cognitive testing every 3 years, incident dementia cases were identified at waves of data collection subsequent to 1991. Of the people contacted to participate in the study, 81% agreed to participate. A total of 1,496 members of the Study of Dementia in Swedish Twins were born from 1886 through 1925 and have complete answers to the leisure activity items from the 1967 questionnaire.
In summary, the present analysis includes all members of the STR born between 1886 and 1914, unless one or both twins were no longer alive, whereas among twins born from 1915 through 1926, only pairs in the SATSA project are included. No dementia assessment was available for other pairs born from 1915 through 1926.
Procedures
Case ascertainment for the Study of Dementia in Swedish Twins used a two-step process: an initial screening and then a comprehensive clinical evaluation that resulted in a consensus diagnosis. The initial screening used the Mini-Mental State Examination (MMSE; Folstein, Folstein, & McHugh, 1975) for those who were visited by the SATSA assessment team and a telephone screening protocol for those who were not visited. Sensitivity of the telephone screening protocol was found to be.86, and specificity was.90 (Gatz et al., 2002). Subsequently, dementia suspects (i.e., those who scored below preestablished norms for cognitive impairment on MMSE or the telephone screening protocol) and their twin partners were evaluated for dementia by an assessment team that consisted of a nurse, a psychologist, and a physician. The protocol used by this team parallels the Consortium to Establish a Registry for Alzheimer's Disease (CERAD; Morris et al., 1989) procedures for physical and neurological evaluations, laboratory tests, informant interview, neuropsychological testing, and neuroimaging. Findings of the assessment team were discussed at a consensus diagnosis conference attended by the clinicians and chaired by a psychologist who had not met the participant. Clinical dementia diagnoses were based on the Diagnostic and Statistical Manual of Mental Disorders (American Psychiatric Association, 1987) criteria for dementia and the National Institute of Neurological and Communicative Disorders and Stroke/Alzheimer's Disease and Related Disorders Association criteria for probable and possible Alzheimer's disease (McKhann et al., 1984). After the National Institute of Neurological Disorders and Stroke–Association Internationale pour la et l'Enseignement en Neurosciences guidelines were published, these criteria were used to determine the presence of vascular dementia (Roman et al., 1993). Twins and their twin partners underwent identical diagnostic procedures. If the partners were deceased, their diagnosis before death was determined through informant interviews and review of medical records, including death certificates. Cases and partners were followed longitudinally every 18 months.
Case ascertainment from the OCTO-Twin sample was essentially parallel procedurally, entailing a review of MMSE scores and neuropsychological testing, with the cognitive battery parallel to that used in the Study of Dementia in Swedish Twins. For those suspected of dementia, the same informant protocol was followed, medical records were reviewed, and a consensus diagnosis was assigned by a physician and a psychologist. The same psychologist chaired the diagnostic conferences for both samples.
Discordant Pairs Sample
A pair was classified as discordant if (a) the partner remained alive and nondemented for at least 5 years after the proband's (i.e., the demented twin's) dementia onset or (b) the partner died during the 5-year follow-up interval but remained alive and nondemented for a time interval specified in a sliding scale for survival (Gatz, Pedersen, Crowe, & Fiske, 2000). For the pairs in which the nondemented partner died within 5 years of dementia onset of the proband, the sliding scale required survival for progressively fewer years as age of onset of the proband's dementia increased. When age of onset was 65 or younger, the partner was required to be alive and free of dementia for the full 5 years. At the upper end of the sliding scale, in pairs in which the proband's age of onset was 75 or older, if the partner died any time after the proband's age of onset and was established to be nondemented at the time of death, the pair was included.
This definition of discordancy reflects several considerations. Other studies have either not used a rule or have used a 3-year interval. For example, Räihä, Kaprio, Koskenvuo, Rajala, and Sourander (1998) simply considered pairs as discordant when a twin had a diagnosis of Alzheimer's disease and the twin partner did not. We chose a more stringent interval based in part on our own survival analysis (Posner, Pedersen, & Gatz, 1999). It indicated that, if a twin partner to a dementia case was also going to become demented, onset was most often within 5 years, after which there was a markedly diminished likelihood. However, with age, there are a growing number of competing causes of mortality. As a result, requiring 5 years of survival at increasingly old ages sharply reduces the number of available twin pairs in the older age groups and possibly misrepresents the association between the predictor and outcome at older ages. Gatz and colleagues (2000) presented analyses comparing different definitions of discordancy and found that the minimum of 5 years of difference in age of onset among twin partners and the sliding scale for survival yielded good sensitivity compared with no rule or the 3-year rule.
The definition of discordancy yielded 107 pairs discordant for dementia. Monozygotic pairs comprised 34% (36 pairs) of the sample, and 66% (71 pairs) were female. Alzheimer's disease accounted for 63% (n = 67) of dementia cases. The average age of dementia onset was 77 (SD = 8.0), and the mean age of the intact partner when last followed was 86 (SD = 5.8). The average interval between age at diagnosis and age at last follow-up or death of the partner was 8.8 years (SD = 5.1). The average age in 1967, when the leisure questionnaire was answered, was 57 (SD = 5.9). Approximately 17% of the twins were reared apart (19 pairs).
Leisure Activities
Members of the STR who were born between 1886 and 1925 were sent a questionnaire in 1967 that included questions on leisure activities. Participants answered 11 items indicating whether or not, before the age of 40, they were regularly involved in each of these activities: reading, listening to the radio or watching television, social visits, cultural activities such as theatre and cinema, hobbies, home and family, clubs and organizations, studies, house and gardening, outdoor activities, and playing sports. Frequency of "yes" responses to each of the individual items is shown in Table 1 for cases and for controls.
|
Results of the factor analysis are summarized in Table 2, showing the highest loading for each item. This analysis yielded three factors, all of which had eigenvalues greater than 1. The "hobbies" item was discarded because it did not load highly onto any one of the factors and showed a low loading on both Factor 1 and Factor 2. The items that loaded on Factor 1 were thought to reflect intellectual–cultural activity, whereas the items that loaded on Factor 2 were thought to reflect self-improvement activities. Factor 3 consisted of variables reflecting domestic activity. Items loading onto the different factors were given unit weights and summed to create a score for each of the factors. The intercorrelations among the three factor scores are shown in Table 3.
|
|
Education
The education variable was scored on a 5-point scale from 1 (compulsory education, or 6 years of school) to 5 (completed university education). On average, participants attained slightly more than the compulsory elementary school education. Level of education was positively correlated with self-improvement (r =.22; p =.02), positively but not significantly correlated with intellectual–cultural activity (r =.11; p =.28), and negatively but not significantly correlated with domestic activity (r = -.09; p =.38).
Statistical Analysis
As a way to evaluate the relative risk of dementia in general, and Alzheimer's disease specifically, associated with participation in leisure activities, matched-pair odds ratios (ORs) were calculated, with the level of education controlled for. Conditional logistic regression was used to determine statistical significance and 95% confidence intervals, using the PROC PHREG procedure for SAS Program Version 8 (SAS Institute, 1999). Some studies have indicated that gender may be associated both with dementia risk (Cummings et al., 1998) and engagement in activities (Hultsch, Hammer, & Small, 1993). For the leisure activities–dementia association within men and women to be examined separately, data were analyzed separately by gender. Stratification by zygosity was used to examine whether there were any effects of including both monozygotic and dizygotic twins in the study sample. Another confounder examined by data stratification was whether the participant was demented at the initial dementia assessment or whether the dementia was discovered in a longitudinal follow-up (i.e., prevalent versus incident cases). Because some twins were raised apart, analyses also compared twins raised apart and together separately. An OR > 1.0 is interpreted as reflecting elevated risk, whereas OR < 1.0 reflects protective effects. Power was calculated by using the formula provided by Schlesselman (1982) for pair-matched studies.
|
RESULTS |
|---|
|
TOPAbstractMethodsResultsDiscussionREFERENCES |
|---|
.05) for all dementias combined as well as for Alzheimer's disease alone. Stratification by gender showed that this protective effect was apparent for women alone but not for men alone.
|
|
|
DISCUSSION |
|---|
|
TOPAbstractMethodsResultsDiscussionREFERENCES |
|---|
The finding that overall level of activity and intellectual–cultural activity were protective for women and not men was unexpected. Whereas we did not find any statistically significant gender differences in this study (possibly because of the lack of power), our results are consistent with the findings of a previous study by Pedersen, Reynolds, and Gatz (1996). In a study of the relationship between education and MMSE scores in a nondemented sample of twins, the authors found that there was greater environmental than genetic mediation of the MMSE–education correlation for women than for men. This suggests that the relative importance of environmental influences on late-life cognitive abilities, such as the effects of engaging in leisure activities, may be more pronounced in women than in men.
A further possibility is that men were engaged in activities that reflect the intellectual–cultural dimension through their occupations more frequently than women. Thus, leisure-time activities would not be as consequential for cognitive functioning in men relative to women. Launer and colleagues (1999) found that women with low education were at greater risk for dementia than men with the same low education. Measuring cognitive change over 7 and 14 years, Schaie's (1984) research found that people who showed the most cognitive decline were widowed women who were never in the workforce and who showed a disengaged lifestyle. Similar to the trends that we observed, these findings suggest an interaction between gender and lifestyle in a broader sense predicting cognitive functioning. In our study, however, gender differences were not significant, and the possibility of a gender–lifestyle interaction remains speculative.
The pattern of our results suggests that, in addition to intellectual–cultural activities, self-improvement activities may be protective against dementia, but low power hindered us from finding a significant association. For domestic activities, there was no trend suggesting an association with dementia risk. Given that participation in self-improvement activities was positively correlated with participation in intellectual–cultural activities, whereas participation in domestic activities was not, our findings seem to support the hypothesis that relatively more intellectual activities would have greater consequence than less intellectual activities. This conclusion is different from the findings from the Friedland and colleagues (2001) and Scarmeas and colleagues (2001) studies, in which greater participation in all categories of activity was related to lower risk. Results from the Paquid (Personnes Agées QUID) study by Fabrigoule and colleagues were more similar to ours in the sense that only a few of the activities by themselves were related to risk of dementia (Fabrigoule et al., 1995). These authors suggested that the significant protective factors all involved attentional control processes. However, we do not feel confident that level of attentional control processes distinguished the three activity factors in our study.
One way that engaging in intellectual–cultural activities could enhance cognitive functioning in later life is by building verbal knowledge. Greater verbal skills could, in turn, lead to better learning abilities, perhaps through more effective use of internal memory aides. Because memory problems are the core feature of dementia, activities that promote maintenance of memory skills would logically be most related to preventing or postponing dementia. One may object that watching television—one of the items loading on the intellectual–cultural factor—does not represent an intellectually challenging task. However, although watching television may not seem to be a very intellectual activity from a present-day perspective, Swedish television during the 1960s and earlier consisted of primarily BBC-type (or PBS) programming. Regardless of how intellectually stimulating television was in Sweden, all of the items loading onto the first factor appeared to involve processing of verbal information. Items loading on the other two factors were less consistent in the extent to which they implicated processing of verbal information.
The lack of detailed information on leisure activities is a main limitation of this study. Unlike Friedland and colleagues (2001), who were able to classify activities in terms of diversity, intensity, and percentage intensity, our study had only yes–no responses to activities. Furthermore, because a variety of differentially complex activities could be subsumed within any one of the items, we could not examine the relationship between complexity of activities and risk of dementia. However, despite this lack of precision in our data, we still found a significant effect of intellectual–cultural activities and overall activity level. In addition, our study used factor analysis to determine which leisure items should go together. Although the naming of these factors was unavoidably somewhat arbitrary, the categorization of our items was empirically based.
Another limitation of this study is the low power for many of the analyses, making null results difficult to interpret. Despite having ascertained a large number of cases of dementia, we found that the number of twin pairs informative for analysis quickly dwindles as a result of being discordant for both dementia and activity level. The finding that ORs were near 1.0 for overall activity in men supports the idea that activity may not be related to risk of dementia in men. However, the lack of significant findings for self-improvement activities may be due mainly to a lack of power rather than lack of a relationship because ORs were not as close to 1.0. The issue of low power necessitates a very cautious interpretation of null results in this study.
It could be argued that a potential limitation of this study is our use of a 5-year age-dependent sliding scale to define dementia discordancy in twins, which is due to the use of less stringent criteria for classification of discordancy among the oldest twin pairs, if the cognitively intact twin has died. To address this issue, we also analyzed our data by using a 5-year rule regardless of age and obtained very similar results to those reported in Table 5. All results that were statistically significant using the 5-year sliding scale were still significant using the 5-year rule with no sliding scale, despite a smaller sample size (88 discordant pairs). This is because the significant findings were slightly stronger (ORs were lower) with the use of the more stringent criteria, offsetting the larger confidence intervals.
A strength of the present study is that there was a 20-year or more time lag between activity data collection and evaluation for dementia. In this sense, our design was prospective. Moreover, because of the time delay between information gathering and follow-up, it is extremely unlikely that early symptoms of dementia affected participation in leisure activities. Although our study had the advantage of a long time delay (avoiding confounding with prediagnosis symptomatology) and use of self-report (avoiding biases inherent in proxy reports), the information gathered was still retrospective. Participants were asked whether they regularly engaged in the leisure activities before the age of 40. Participants were aged 42 to 68 at the time of data collection in 1967. Although retrospective, answers nonetheless could not have been biased by the outcome—dementia—the onset of which was 20 or more years later.
Greater intellectual ability has been found to affect the types of activities in which people participate (Hultsch, Hertzog, Small, & Dixon, 1999; Schooler & Mulatu, 2001) and has been associated with lower risk of dementia (Whalley et al., 2000) and better cognitive functioning in late life (Plassman et al., 1995). Unfortunately, this study was not able to provide a final answer to the lingering question of whether better cognitive abilities lead to engagement in more intellectually stimulating activities or vice versa. Hultsch and colleagues (1999) stated that "this conundrum can probably never be completely resolved" (p. 246). In our study, controlling for education represented the only available method to control for differences in cognitive abilities. However, the finding that trends were similar in a comparison of monozygotic with dizygotic pairs gives us more confidence that the activities were indeed protective beyond representing greater intellectual ability to begin with. In line with other studies (Fabrigoule et al., 1995; Friedland et al., 2001; Kondo et al., 1994; Scarmeas et al., 2001; Wilson et al., 2002) and conclusions drawn by other researchers, our study suggests that, at worst, there may be no effect of leisure activities on risk of dementia and that, at best, engaging in certain types of leisure activities may be protective.
|
Acknowledgments |
|---|
Received for publication April 16, 2002. Accepted for publication March 28, 2003.
|
REFERENCES |
|---|
|
TOPAbstractMethodsResultsDiscussionREFERENCES |
|---|
This article has been cited by other articles:
 |
|
 |
|
  E. X. Albuquerque, E. F. R. Pereira, M. Alkondon, and S. W. Rogers Mammalian Nicotinic Acetylcholine Receptors: From Structure to Function Physiol Rev, January 1, 2009; 89(1): 73 - 120. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. Kroger, R. Andel, J. Lindsay, Z. Benounissa, R. Verreault, and D. Laurin Is Complexity of Work Associated with Risk of Dementia?: The Canadian Study of Health and Aging Am. J. Epidemiol., April 1, 2008; 167(7): 820 - 830. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. S. Wilson, P. A. Scherr, J. A. Schneider, Y. Tang, and D. A. Bennett Relation of cognitive activity to risk of developing Alzheimer disease Neurology, November 13, 2007; 69(20): 1911 - 1920. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. A. M. Bielak, T. F. Hughes, B. J. Small, and R. A. Dixon It's Never Too Late to Engage in Lifestyle Activities: Significant Concurrent but not Change Relationships Between Lifestyle Activities and Cognitive Speed J. Gerontol. B. Psychol. Sci. Soc. Sci., November 1, 2007; 62(6): P331 - P339. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. C. Park, A. H. Gutchess, M. L. Meade, and E. A. L. Stine-Morrow Improving Cognitive Function in Older Adults: Nontraditional Approaches J. Gerontol. B. Psychol. Sci. Soc. Sci., June 1, 2007; 62(suppl_Special_Issue_1): 45 - 52. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
B. M. van Gelder, M. Tijhuis, S. Kalmijn, S. Giampaoli, A. Nissinen, and D. Kromhout Marital Status and Living Situation During a 5-Year Period Are Associated With a Subsequent 10-Year Cognitive Decline in Older Men: The FINE Study. J. Gerontol. B. Psychol. Sci. Soc. Sci., July 1, 2006; 61(4): P213 - P219. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. Rundek and D. A. Bennett Cognitive leisure activities, but not watching TV, for future brain benefits Neurology, March 28, 2006; 66(6): 794 - 795. [Full Text] [PDF]
|
|
|
|
|
|
J. S. Saczynski, L. A. Pfeifer, K. Masaki, E. S. C. Korf, D. Laurin, L. White, and L. J. Launer The Effect of Social Engagement on Incident Dementia: The Honolulu-Asia Aging Study Am. J. Epidemiol., March 1, 2006; 163(5): 433 - 440. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
F. Beland, M.-V. Zunzunegui, B. Alvarado, A. Otero, and T. del Ser Trajectories of Cognitive Decline and Social Relations J. Gerontol. B. Psychol. Sci. Soc. Sci., November 1, 2005; 60(6): P320 - P330. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. Andel, M. Crowe, N. L. Pedersen, J. Mortimer, E. Crimmins, B. Johansson, and M. Gatz Complexity of Work and Risk of Alzheimer's Disease: A Population-Based Study of Swedish Twins J. Gerontol. B. Psychol. Sci. Soc. Sci., September 1, 2005; 60(5): P251 - P258. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||
| HOME | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
