Journal of Clinical and Experimental Neuropsychology 1380-3395/03/2505-625$16.

00
2003, Vol. 25, No. 5, pp. 625–633 # Swets & Zeitlinger

Cognitive Reserve and Lifestyle

Nikolaos Scarmeas and Yaakov Stern
Cognitive Neuroscience Division, Department of Neurology, Taub Institute for Research in Alzheimer’s Disease
and the Aging Brain, and College of Physicians and Surgeons, Columbia University, New York, NY, USA

ABSTRACT

The concept of cognitive reserve (CR) suggests that innate intelligence or aspects of life experience like
educational or occupational attainments may supply reserve, in the form of a set of skills or repertoires that
allows some people to cope with progressing Alzheimer’s disease (AD) pathology better than others. There is
epidemiological evidence that lifestyle characterized by engagement in leisure activities of intellectual and
social nature is associated with slower cognitive decline in healthy elderly and may reduce the risk of
incident dementia. There is also evidence from functional imaging studies that subjects engaging in such
leisure activities can clinically tolerate more AD pathology. It is possible that aspects of life experience like
engagement in leisure activities may result in functionally more efficient cognitive networks and therefore
provide a CR that delays the onset of clinical manifestations of dementia.

The CR hypothesis suggests that there are indivi- supply reserve, in the form of a set of skills or
dual differences in the ability to cope with AD repertoires that allows some people to cope with
pathology (Stern, 2002). For example, Katzman pathology better than others.
et al. (1989) described cases of cognitively nor- Epidemiological data supporting the CR hy-
mal, elderly women who were discovered to have pothesis include observations that lower educa-
advanced AD pathology in their brains at death. tional and occupational attainment is associated
They speculated these women did not express the with increased risk for incident dementia (Stern
clinical features of AD because their brains were et al., 1994). Similarly, lower linguistic ability (as
larger than average. About 25% of subjects who expressed by idea density and grammatical
during autopsy fulfill pathologic criteria for AD complexity) in early life and childhood mental
and were assessed and followed in well-charac- ability scores are strong predictors of poor
terized cohorts were clinically intact during life cognitive function and dementia in late life
(Ince, 2001). Similarly, most clinicians are aware (Snowdon et al., 1997; Whalley et al., 2000).
of the fact that a stroke of a given magnitude can This is consistent with the prediction that people
produce profound impairment in 1 patient and with more reserve can cope with advancing AD
while having minimal effect on another. Some- pathology longer before it is expressed clinically.
thing must account for the disjunction between In addition it has been shown that AD patients
the degree of brain damage and its outcome, and with higher educational and occupational attain-
the concept of reserve has been proposed to serve ment have more rapid cognitive decline than those
this purpose. with lower attainment, consistent with the idea
Innate intelligence or aspects of life experience that, at any level of clinical severity, the under-
like educational or occupational attainment may lying AD pathology is more advanced in patients

Address correspondence to: Nikolaos Scarmeas, Columbia Presbyterian Medical Center, P&S Mailbox 16, PH 19th
Floor, 630 West 168th Street, New York, NY 10032, USA. Tel.: þ 1-212-305-9194. E-mail: [email protected]
Accepted for publication: October 29, 2002.
626 NIKOLAOS SCARMEAS & YAAKOV STERN

with more CR (Stern, Tang, Denaro, & Mayeux, social and physical activities? Longitudinal data
1995). can offer a partial resolution of this conundrum.
At least three studies have used structural
equation modeling in longitudinal data to address
LEISURE ACTIVITIES this question. Schooler and Mulatu (2001),
reported that initial high levels of intellectual
Factors other than IQ, education and occupation functioning leads to high levels of environmental
might also provide reserve and influence the complexity, which in turn raises the levels of
incidence of AD. It has been theorized that intellectual functioning over a 20-year period.
changes in everyday experiences and activity Gold et al. (1995) reported that individuals with
patterns may result in disuse and consequent higher levels of intellectual ability, education and
atrophy of cognitive processes and skills (a view socioeconomic status are more likely to develop
captured in the adage ‘‘use it or loose it’’; an engaged lifestyle, which in turn contributes to
Salthouse, 1991). Taking into account the con- the maintenance of verbal intelligence in later
siderable plasticity of cognitive abilities of older life. Similar observations were reported in a 250
adults, one might predict that deliberate practice individual sample tested three times over 6 years:
of such skills would at least result in stable intellectually engaging activities seemed to buffer
performance or may even reverse age-related against decline in cognitive functioning (Hultsch
changes. Does the stimulation provided by typical et al., 1999). Nevertheless, an alternative model
everyday activities facilitate the maintenance and in this study suggested that findings were also
improvement of general cognitive skills via expo- consistent with the hypothesis that high-ability
sure to cognitive training (Hultsch, Hertzog, individuals lead intellectually active lives until
Small, & Dixon, 1999)? In other words, could cognitive decline in old age limits their activities.
everyday experience affect cognition in a manner In another study, individuals with high socio-
that is analogous to physical exercise for muscu- economic status who were fully engaged with
losceletal and cardiovascular functions? their environment had the least intellectual
decline over 7 and 14 years (Schaie, 1984,
1996). In this study it was widowed women who
HEALTHY ADULTS had never been in the workforce and who
exhibited a disengaged lifestyle that exhibited
Many studies have investigated the association the greatest decline. Finally, Arbuckle, Gold,
between level of participation in activities and Andres, Schwartzman, and Chaikelson (1992)
performance on various cognitive tasks in healthy reported that participation in intellectual activities
adults (Arbuckle, Gold, & Andres, 1986; was related to maintenance of intellectual perfor-
Christensen et al., 1996; Craik, Byrd, & Swanson, mance in a sample of 2nd World War veterans
1987; Erber & Szuchman, 1996; Hill, Wahlin, tested twice over a 40-year period.
Winblad, & Backman, 1995; Hultsch, Hammer, &
Small, 1993; Luszcz, Bryan, & Kent, 1997; van
Boxtel, Langerak, Houx, & Jolles, 1996). In ALZHEIMER’S DISEASE
general, these studies have reported that there is
a positive association between participation in Although educational and occupational attain-
intellectual, social and physical activities and ments have been extensively studied, there are
performance on a wide range of cognitive tasks. very few reports examining the influence of
Nevertheless, the lack of temporal depth of cross- socially and intellectually engaged lifestyle to
sectional studies raises concerns with issues of dementia. One case control study of AD in
causal directionality. Does participation in stimu- Japan (Kondo, Niino, & Shido, 1994; 60 cases),
lating activities promote cognitive performance or reported that cases were significantly less active
is it that better performing cognitively capable in various use of leisure time, hobbies and psy-
subjects tend to participate in more intellectual, chosocial behaviors. Another case-control study
 COGNITIVE RESERVE AND LIFESTYLE 627

examined the presence of nonoccupational activ- deficiencies or disabilities. When the comparison
ities during midlife in 193 subjects with possible group was subjects who were not engaged in these
and probable AD (Friedland et al., 2001). When two activities for other reasons, no significant
activity patterns were classified into intellectual, influence was noted. Therefore, the reasons for
passive and physical, cases were less likely to engaging in leisure activities, rather than the
have participated in intellectual activities. leisure activities per se, seemed to affect incident
There have been only few prospective lon- dementia in this study.
gitudinal studies examining the influence of There have been at least three large prospec-
socially and intellectually engaged lifestyle to tive studies that demonstrated a clear association
incident dementia. In a survey sample of 422 between engaged lifestyle and incident dementia.
elderly subjects the relation of various indicators In one of them, social ties (spouse, living
of socio-economic status to incident dementia arrangements, contacts with friends and relatives,
was investigated (Bickel & Cooper, 1994). Only confidant reciprocity and group membership) and
poor quality living accommodations were asso- participation in productive activities (helping
ciated with increased risk of incident dementia, others with daily tasks, paid work and volunteer
while indicators of social isolation such as low work) were assessed for 2486 community-dwell-
frequency of social contacts within and outside ing Japanese-American men (free of dementia at
the family circle, low standard of social support baseline) who were followed for 3 years (Balfour,
and living in single person household did not Masaki, White, & Launer, 2001). Both social
prove to be significant. engagement and productive activities were sig-
In another cohort, the average minute per day nificantly and inversely associated with risk of
participation in cognitively challenging ‘‘com- incident dementia. With each additional social tie,
plex’’ (i.e., playing a musical instrument, handi- the OR for dementia decreased by 16% and with
crafts) and basic ‘‘simple’’ (dressing, eating) each additional productive activity, the OR for
activities was assessed for 1877 subjects (Zabar, dementia decreased by 72% and the effect of the
Corrada, Fozard, Costa, & Kawas, 1996). Regular two scales was synergistic. This study is available
participation in complex (relative to simple) only in an abstract form.
activities did reduce the risk of developing We recently reported the results of a study
dementia within 2 years but not within 5, 10 or involving a total of 1772 nondemented individuals
15 years, suggesting that the effect near the time aged 65 years or older, living in Northern
of diagnosis may reflect changes in activities due Manhattan New York who were identified and fol-
to early dementia. Unfortunately, this study has lowed longitudinally in a community based cohort
been published only in an abstract form and no incidence study for up to 7 years (mean 2.9 years;
details are available. Scarmeas, Levy, Tang, Manly, & Stern, 2001). In
Another study evaluated social and leisure the initial visit, an interview elicited self-reported
activity data in 2040 nondemented elderly com- participation in a variety of leisure activities of
munity residents from Gironde (France) and intellectual (reading magazines or newspapers or
recorded incident dementia on follow-up visits books, playing cards or games or bingo, going to
(Fabrigoule et al., 1995). Traveling, doing odd classes etc.), social (visiting or being visited by
jobs and knitting were associated with lower risk friends or relatives, etc.) and even physical (walk-
of incident dementia when occupational status ing for pleasure or excursion, physical condition-
was controlled for. However, the analyses were ing, etc.) nature. Figure 1 presents the cumulative
not controlled for either ethnic group, gender, risk for developing dementia in the high and low
educational attainment, cerebrovascular risk fac- leisure activity groups, as calculated from survival
tors or depressive symptomatology. In addition, analysis models. Even when factors like ethnic
for doing odd jobs and knitting, the effect was group, education and occupation were controlled
significant only when subjects who did the for, subjects with high leisure activity had 38%
activities without difficulty were compared to less risk of developing dementia. The risk of
subjects who did not do them because of incident dementia was reduced by approximately
628 NIKOLAOS SCARMEAS & YAAKOV STERN

Fig. 1. Survival curve based on Cox analysis comparing cumulative dementia incidence in subjects with high and
low leisure activities in three age groups of the cohort (Scarmeas et al., 2001).

12% for each additional leisure activity adopted. cognitive performance, health limitations inter-
The effect of leisure activities on incident fering with social activities, cerebrovascular
dementia was still present even when baseline disease and depression were considered.
 COGNITIVE RESERVE AND LIFESTYLE 629

In another prospective study, frequency of advanced pathology). It has been shown that pa-
participation in common cognitive activities (i.e., tients with higher educational (Stern, Alexander,
reading a newspaper, magazine, books) was Prohovnik, & Mayeux, 1992), or occupational
assessed at baseline for 801 elderly Catholic (Stern, Alexander et al., 1995) attainment, as well
nuns, priests and brothers without dementia as those with higher premorbid IQ (Alexander
(Wilson et al., 2002). During a mean follow-up et al., 1997) have more prominent flow deficits
of 4.5 years, 1-point increase in the cognitive (and hence more pathology) when controlling for
activity score was associated with a 33% reduc- clinical severity. These observations support the
tion in the risk for AD. Additionally, engagement prediction that individuals with more reserve can
in cognitive activities was also associated with tolerate more pathology.
slower rates of cognitive decline. The results held Results from a recent study invoking leisure
even when many potential confounders were activities seem to parallel those for education,
controlled for and even when subjects with occupation and IQ. We evaluated leisure (intel-
memory impairment at baseline evaluation were lectual, social, and physical) activities in 9
excluded from the analyses. patients with early AD and 16 healthy elderly
In contrast to previous studies that were who underwent brain H2 15 O PET (Scarmeas et al.,
investigating risk for incident dementia, another in press). In a voxel-wise multiple regression
study by the same group has explored the rate of analysis which controlled for clinical severity,
cognitive decline in relation to premorbid reading there was a negative correlation between leisure
activity in subjects who had already manifested activity score and CBF. When education, esti-
the disease. In a cohort of 410 persons with AD mates of premorbid IQ or both were added as
followed for a 4-year period higher levels of covariates in the same model, higher leisure
premorbid reading activity was associated with activities score was still associated with more
more rapid decline in the global cognitive and prominent CBF deficits. These results corroborate
verbal measures (Wilson et al., 2000). Although the hypothesis that at any given level of clinical
counterintuitive at first glance, the results are disease severity, more severe AD pathology exists
consistent with the hypothesis that intellectual in the patients with high leisure activities, even
activities may enhance brain’s reserve capacity when education and IQ are taken into account.
and that at any level of clinical severity, the
underlying AD pathology is more advanced in
patients with more CR. Therefore, although NATURE OF ACTIVITIES
subjects with higher CR (compared to ones with
lower CR) may manifest dementia symptomatol- Although it makes intuitive sense that only cog-
ogy later in life, they may decline faster after nitively challenging activities might be related to
dementia onset. These results for reading activity risk for dementia there is also evidence for pro-
parallel previous similar ones for education and tection even for noncognitive activities. In our
occupation (Stern, Tang et al., 1995). study, when the leisure items were grouped into
physical, social and intellectual factors, although
the intellectual factor was associated with lowest
IMAGING STUDIES risk of incident dementia all three factors retained
their significant effect in the analyses (Scarmeas
Physiologic data from functional imaging studies et al., 2001).
have served as an indirect affirmation of the Epidemiological evidence that physical exer-
reserve hypothesis. Given that the regional corre- cise may delay cognitive impairment is equivocal.
lation between PET cerebral blood flow (CBF) While high levels of physical activity were
deficits and histologically confirmed postmortem associated with reduced risk of dementia in at
dementia changes is well-established, CBF has least four prospective studies (Laurin, Verreault,
been used as a indirect-surrogate measure of Lindsay, MacPherson, & Rockwood, 2001; Li
disease pathology (lower CBF indicating more et al., 1991; Scarmeas et al., 2001; Yoshitake et al.,
630 NIKOLAOS SCARMEAS & YAAKOV STERN

1995), no effect of exercise on dementia and Another explanation of the findings could be
cognitive impairment risk was reported from that borderline dementia subjects might have
other cohorts (Broe et al., 1998; Wilson et al., lower leisure activity as a result of early disease.
2002). Additionally, there is basic research The consequence of such a premise would be that
evidence that environmental enrichment in the low leisure activity as recorded in epidemiologi-
form of voluntary wheel running is associated cal studies represents a manifestation of early
with enhanced neurogenesis in the adult mouse dementia rather than a premorbid risk factor per
dentate gyrus (van Praag, Kempermann, & Gage, se. Some studies attempted to partially address
1999). It has also been shown that physical this possibility by excluding from the analyses
activity sustains cerebral blood flow (Rogers, subjects with memory impairment at baseline
Meyer, & Mortel, 1990) and it may improve evaluation or by considering baseline cognitive
aerobic capacity and cerebral nutrient supply performance in the analyses (Scarmeas et al.,
(Dustman et al., 1984; Spirduso, 1980). There- 2001; Wilson et al., 2002). Although the protec-
fore, although it is conceivable that physical tive effect for leisure activities remained
activity may merely be a nonspecific marker of unchanged in these studies, the activities were
good health indirectly related to dementia (or recorded only a few years before dementia
even not related to dementia at all), it is also incidence. The longer the interval between
possible that it has a direct physiological assessment of such activities and dementia
association with brain disease. diagnosis, the more confident one can be about
the temporality and causality of the relation.
Overall, given the current literature, the scenario
IS THE ASSOCIATION CAUSAL? of lifestyle being affected by subtle, incipient
cerebral disease cannot be completely excluded.
The association between engaged lifestyle and
dementia risk could be either mediated or con-
IF THE ASSOCIATION IS TRULY
founded by abilities like IQ or education. If this is
CAUSAL, HOW MAY IT TAKE EFFECT?
the case then it could be that either IQ or educa-
tion represent the true causal links with dementia
Leisure activities may mediate protection via
or that subjects with higher IQ or education tend
many different ways – various versions of the
to adopt lifestyles which themselves causally
CR hypothesis.
reduce the risk of dementia (such as exercise,
diet, etc.). Nevertheless, in studies where educa- 1. Passive or hardware: Bigger brains tolerate
tion and occupation (Scarmeas et al., 2001) or more loss before exhibiting impaired function
education and IQ (Scarmeas et al., in press) were because of higher number of healthy synapses
controlled for, the association between leisure or neurons resulting in increased number of
activities and dementia risk was still there. remaining available ones when a certain per-
It is also possible that both education-occupa- centage of them is affected by a pathologic
tion-IQ and engagement in stimulating vocational process (Katzman et al., 1988). Intellectually
activities are markers of innate capacities. Innate and socially engaged lifestyle may increase
might refer to either genetic background or early synaptic density in neocortical association
life developmental factors or a combination of cortex (on the basis of stimulation (Katzman,
them. These innate capacities might in turn lead to 1993)) which may result in more efficient
higher levels of education, more engagement in cognitive function of unaffected neurons that
stimulating vocational activities and lower risk of might be able to compensate for loss of func-
cognitive decline. tion of affected brain areas.
It can also be that there is another yet not 2. Active or software: More efficient use of the
identified causal factor-characteristic that con- same brain networks. Even though the number
founds or mediates the inverse association of neurons or synapses might be the same,
between engaged lifestyle and incident dementia. enhanced synaptic activity or more efficient
 COGNITIVE RESERVE AND LIFESTYLE 631

circuits of synaptic connectivity might exist in brain to mediate a task by making them more
subjects exerting more leisure activities. efficient or resilient in the face of brain pathology
3. Active or software: More efficient use of or by recruitment of alternate networks. Inter-
alternative brain networks, that is, more effi- individual differences in lifestyle may partially
cient ability to shift operations to alternate mediate the relationship between brain pathology
circuits. As a concrete example, a trained and the clinical manifestation of AD and engaged
mathematician or somebody with lifelong lifestyle may supply a reserve that allows an
engagement in mathematical training, might individual to cope longer before AD is clinically
be able to solve a mathematics problem many expressed.
different ways, while a less experienced indi-
vidual might have only one possible solution
strategy available. The mathematician would REFERENCES
have more flexibility in solving the problem if
any particular solution strategy was precluded. Alexander, G.E., Furey, M.L., Grady, C.L., Pietrini, P.,
This built in redundancy would permit greater Brady, D.R., Mentis, M.J., & Schapiro, M.B. (1997).
resilience in the face of brain damage. Association of premorbid intellectual function with
cerebral metabolism in Alzheimer’s disease: Impli-
cations for the cognitive reserve hypothesis. Amer-
Finally, it is conceivable that factors that affect ican Journal of Psychiatry, 154, 165–172.
CR may even result in hindering the development Arbuckle, T.Y., Gold, D., & Andres, D. (1986).
of the disease pathology per se. Involvement in Cognitive functioning of older people in relation
challenging avocations during life may even to social and personality variables. Psychology and
decrease neurodegeneration. This could be Aging, 1, 55–62.
mediated via a variety of underlying neurobiolo- Arbuckle, T.Y., Gold, D.P., Andres, D., Schwartzman,
A., & Chaikelson, J. (1992). The role of psychoso-
gical processes such as enhanced chronic neuro-
cial context, age, and intelligence in memory
nal activation associated with increased brain performance of older men. Psychology and Aging,
work, increased regional cerebral blood flow, and 7, 25–36.
increased glucose and oxygen metabolism Balfour, J.L., Masaki, K., White, L., & Launer, L.J.
(Friedland, 1993) or even via higher ability for (2001). The effect of social engagement and
generation of new neurons into adulthood productive activity on incident dementia:
(Eriksson et al., 1998; Johansson, Svensson, The Honolulu Asia Aging Study. Neurology,
56(Suppl.), A239.
Wallstedt, Janson, & Frisen, 1999; Kukekov
Bickel, H., & Cooper, B. (1994). Incidence and relative
et al., 1999) that might help to make up cell loses risk of dementia in an urban elderly population:
with cell gains. Findings of a prospective field study. Psychological
The hypothesis of lifestyle modifications Medicine, 24, 179–192.
affecting brain disease outcome becomes even Broe, G.A., Creasey, H., Jorm, A.F., Bennett, H.P.,
more exciting in the face of increasing animal Casey, B., Waite, L.M., Grayson, D.A., & Cullen, J.
research literature about neuronal plasticity and its (1998). Health habits and risk of cognitive impair-
ment and dementia in old age: A prospective study
association with learning, and physically and
on the effects of exercise, smoking and alcohol
socially enriched environment. Exposure to an consumption. Australian and New Zealand Journal
enriched environment, defined as a combination of of Public Health, 22, 621–623.
more opportunities for physical activity, learning Christensen, H., Korten, A., Jorm, A.F., Henderson,
and social interaction, produces not only a host of A.S., Scott, R., & Mackinnon, A.J. (1996). Activity
structural and functional changes in the brain but levels and cognitive functioning in an elderly
also influences the rate of neurogenesis in adult community sample. Age and Ageing, 25, 72–80.
Craik, F.I., Byrd, M., & Swanson, J.M. (1987). Patterns
and senescent animal models (Kempermann,
of memory loss in three elderly samples. Psychol-
Kuhn, & Gage, 1997a, 1997b; van Praag et al., ogy and Aging, 2, 79–86.
1999). Dustman, R.E., Ruhling, R.O., Russell, E.M., Shearer,
Aspects of life experience like leisure activ- D.E., Bonekat, H.W., Shigeoka, J.W., Wood, J.S., &
ities could modify the paradigms used by the Bradford, D.C. (1984). Aerobic exercise training
632 NIKOLAOS SCARMEAS & YAAKOV STERN

and improved neuropsychological function of older Katzman, R., Aronson, M., Fuld, P., Kawas, C., Brown,
individuals. Neurobiology of Aging, 5, 35–42. T., Morgenstern, H., Frishman, W., Gidez, L., Eder,
Erber, J.T., & Szuchman, L.T. (1996). Memory H., & Ooi, W.L. (1989). Development of dementing
performance in relation to age, verbal ability, and illnesses in an 80-year-old volunteer cohort. Annals
activity. Experimental Aging Research, 22, 59–72. of Neurology, 25, 317–324.
Eriksson, P.S., Perfilieva, E., Bjork-Eriksson, T., Katzman, R., Terry, R., DeTeresa, R., Brown, T.,
Alborn, A.M., Nordborg, C., Peterson, D.A., & Davies, P., Fuld, P., Renbing, X., & Peck, A. (1988).
Gage, F.H. (1998). Neurogenesis in the adult human Clinical, pathological, and neurochemical changes
hippocampus. Natural Medicines, 4, 1313–1317. in dementia: A subgroup with preserved mental
Fabrigoule, C., Letenneur, L., Dartigues, J.F., Zarrouk, status and numerous neocortical plaques. Annals of
M., Commenges, D., & Barberger-Gateau, P. Neurology, 23, 138–144.
(1995). Social and leisure activities and risk of Kempermann, G., Kuhn, H.G., & Gage, F.H. (1997a).
dementia: A prospective longitudinal study. Journal Genetic influence on neurogenesis in the dentate
of the American Geriatrics Society, 43, 485–490. gyrus of adult mice. Proceedings of the National
Friedland, R.P. (1993). Epidemiology, education, and Academy of Sciences of the United States of
the ecology of Alzheimer’s disease. Neurology, 43, America, 94, 10409–10414.
246–249. Kempermann, G., Kuhn, H.G., & Gage, F.H. (1997b).
Friedland, R.P., Fritsch, T., Smyth, K.A., Koss, E., More hippocampal neurons in adult mice living in
Lerner, A.J., Chen, C.H., Petot, G.J., & Debanne, an enriched environment. Nature, 386, 493–495.
S.M. (2001). Patients with Alzheimer’s disease have Kondo, K., Niino, M., & Shido, K. (1994). A case-
reduced activities in midlife compared with healthy control study of Alzheimer’s disease in Japan –
control-group members. Proceedings of the significance of life-styles. Dementia, 5, 314–326.
National Academy of Sciences of the United States Kukekov, V.G., Laywell, E.D., Suslov, O., Davies, K.,
of America, 98, 3440–3445. Scheffler, B., Thomas, L.B., O’Brien, T.F.,
Gold, D.P., Andres, D., Etezadi, J., Arbuckle, T., Kusakabe, M., & Steindler, D.A. (1999). Multi-
Schwartzman, A., & Chaikelson, J. (1995). Struc- potent stem/progenitor cells with similar properties
tural equation model of intellectual change and arise from two neurogenic regions of adult human
continuity and predictors of intelligence in older brain. Experimental Neurology, 156, 333–344.
men [published erratum appears in Psychology and Laurin, D., Verreault, R., Lindsay, J., MacPherson, K.,
Aging 1998; 13(3): 434]. Psychology and Aging, 10, & Rockwood, K. (2001). Physical activity and risk
294–303. of cognitive impairment and dementia in elderly
Hill, R.D., Wahlin, A., Winblad, B., & Backman, L. persons. Archives of Neurology, 58, 498–504.
(1995). The role of demographic and life style Li, G., Shen, Y.C., Chen, C.H., Zhau, Y.W., Li, S.R., &
variables in utilizing cognitive support for episodic Lu, M. (1991). A three-year follow-up study of age-
remembering among very old adults. Journal of related dementia in an urban area of Beijing. Acta
Gerontology B Psychological Science and Social Psychiatrica Scandinavica, 83, 99–104.
Science, 50, 219–227. Luszcz, M.A., Bryan, J., & Kent, P. (1997). Predicting
Hultsch, D.F., Hammer, M., & Small, B.J. (1993). Age episodic memory performance of very old men and
differences in cognitive performance in later life: women: Contributions from age, depression, activ-
Relationships to self-reported health and activity ity, cognitive ability, and speed. Psychology and
life style. Journal of Gerontology, 48, 1–11. Aging, 12, 340–351.
Hultsch, D.F., Hertzog, C., Small, B.J., & Dixon, R.A. Rogers, R.L., Meyer, J.S., & Mortel, K.F. (1990). After
(1999). Use it or lose it: Engaged lifestyle as a buffer reaching retirement age physical activity sustains
of cognitive decline in aging? Psychology and cerebral perfusion and cognition. Journal of the
Aging, 14, 245–263. American Geriatrics Society, 38, 123–128.
Ince, P. (2001). Pathological correlates of late-onset Salthouse, T. (1991). Theoretical perspectives on cog-
dementia in a multicenter community-based nitive aging. Hillsdale, NJ: L. Erlbaum Associates.
population in England and Wales. Lancet, 357, Scarmeas, N., Levy, G., Tang, M., Manly, J., & Stern,
169–175. Y. (2001). Influence of leisure activity on the
Johansson, C.B., Svensson, M., Wallstedt, L., Janson, incidence of Alzheimer’s disease. Neurology, 57,
A.M., & Frisen, J. (1999). Neural stem cells in the 2236–2242.
adult human brain. Experimental Cell Research, Scarmeas, N., Zarahn, E., Anderson, K.E., Habeck, C.,
253, 733–736. Hilton, J., Flynn, J., Marder, K., Bell, K., Sackeim,
Katzman, R. (1993). Education and the prevalence of H., Van Heertum, R., Moeller, J.R., & Stern, Y.
dementia and Alzheimer’s disease. Neurology, 43, (2002). Association of life activities with cerebral
13–20. blood flow in Alzheimer’s disease: Implications for
 COGNITIVE RESERVE AND LIFESTYLE 633

the cognitive reserve hypothesis. Archives of Stern, Y., Tang, M.X., Denaro, J., & Mayeux, R. (1995).
Neurology, 60, 359–365. Increased risk of mortality in Alzheimer’s disease
Schaie, K. (1984). Midlife influences upon intellectual patients with more advanced educational and
functioning in old age. International Journal of occupational attainment. Annals of Neurology, 37,
Behavioral Development, 7, 463–478. 590–595.
Schaie, K.W. (1996). Intellectual development in adult- van Boxtel, M.P., Langerak, K., Houx, P.J., & Jolles, J.
hood: The Seattle longitudinal study. New York: (1996). Self-reported physical activity, subjective
Cambridge University Press. health, and cognitive performance in older adults.
Schooler, C., & Mulatu, M.S. (2001). The reciprocal Experimental Aging Research, 22, 363–379.
effects of leisure time activities and intellectual van Praag, H., Kempermann, G., & Gage, F.H. (1999).
functioning in older people: A longitudinal analysis. Running increases cell proliferation and neuro-
Psychology and Aging, 16, 466–482. genesis in the adult mouse dentate gyrus. Nature
Snowdon, D.A., Greiner, L.H., Mortimer, J.A., Riley, Neuroscience, 2, 266–270.
K.P., Greiner, P.A., & Markesbery, W.R. (1997). Whalley, L.J., Starr, J.M., Athawes, R., Hunter, D.,
Brain infarction and the clinical expression of Pattie, A., & Deary, I.J. (2000). Childhood mental
Alzheimer disease. The Nun Study. Journal of the ability and dementia. Neurology, 55, 1455–1459.
American Medical Association, 277, 813–817. Wilson, R.S., Bennett, D.A., Gilley, D.W., Beckett,
Spirduso, W.W. (1980). Physical fitness, aging, and L.A., Barnes, L.L., & Evans, D.A. (2000). Pre-
psychomotor speed: A review. Journal of Gerontol- morbid reading activity and patterns of cognitive
ogy, 35, 850–865. decline in alzheimer disease. Archives of Neurology,
Stern, Y. (2002). What is cognitive reserve? Theory and 57, 1718–1723.
research application of the reserve concept. Journal Wilson, R.S., Mendes de Leon, C.F., Barnes, L.,
of the International Neuropsychological Society, 8, Schneider, J.A., Bienias, J.L., Evans, D.A., David,
448–460. A., & Bennett, D.A. (2002). Participation in
Stern, Y., Alexander, G.E., Prohovnik, I., & Mayeux, R. cognitively stimulating activities and risk of inci-
(1992). Inverse relationship between education and dent alzheimer disease. Journal of the American
parietotemporal perfusion deficit in Alzheimer’s Medical Association, 287, 742–748.
disease. Annals of Neurology, 32, 371–375. Yoshitake, T., Kiyohara, Y., Kato, I., Ohmura, T.,
Stern, Y., Alexander, G.E., Prohovnik, I., Stricks, L., Iwamoto, H., Nakayama, K., Ohmori, S.,
Link, B., Lennon, M.C., & Mayeux, R. (1995). Nomiyama, K., Kawano, H., & Ueda, K. (1995).
Relationship between lifetime occupation and Incidence and risk factors of vascular dementia and
parietal flow: Implications for a reserve against Alzheimer’s disease in a defined elderly Japanese
Alzheimer’s disease pathology. Neurology, 45, population: The Hisayama Study. Neurology, 45,
55–60. 1161–1168.
Stern, Y., Gurland, B., Tatemichi, T.K., Tang, M.X., Zabar, Y., Corrada, M., Fozard, J., Costa, P., & Kawas,
Wilder, D., & Mayeux, R. (1994). Influence of C. (1996). Does frequent participation in
education and occupation on the incidence of cognitively demanding leisure activities reduce the
Alzheimer’s disease. Journal of the American risk of developing dementia? Neurology, 46(Suppl.),
Medial Association, 271, 1004–1010. A435.