Personal View

Cognitive reserve in ageing and Alzheimer’s disease

Yaakov Stern

Lancet Neurol 2012; 11: 1006–12 The concept of cognitive reserve provides an explanation for differences between individuals in susceptibility to age-
Cognitive Neuroscience related brain changes or pathology related to Alzheimer’s disease, whereby some people can tolerate more of these
Division, Department of changes than others and maintain function. Epidemiological studies suggest that lifelong experiences, including
Neurology and Taub Institute,
Columbia University College of
educational and occupational attainment, and leisure activities in later life, can increase this reserve. For example, the
Physicians and Surgeons, New risk of developing Alzheimer’s disease is reduced in individuals with higher educational or occupational attainment.
York, NY, USA (Y Stern PhD) Reserve can conveniently be divided into two types: brain reserve, which refers to differences in the brain structure
Correspondence to: that may increase tolerance to pathology, and cognitive reserve, which refers to differences between individuals in
Dr Yaakov Stern, 630 W 168th how tasks are performed that might enable some people to be more resilient to brain changes than others. Greater
Street, P&S Box 16, New York,
NY 10032, USA
understanding of the concept of cognitive reserve could lead to interventions to slow cognitive ageing or reduce the
ys11@columbia.edu risk of dementia.

Introduction thus accounting for the discontinuity. A convenient,

The possibility of a connection between life experience although somewhat artificial, way to view cognitive
and the prevalence of dementia has long been discussed. reserve is to separate it into two main features: brain
In 1981, Gurland1 wrote “It is still an open matter whether reserve and cognitive reserve.
there is an important sociocultural contribution to the The original concept of brain reserve was quantitative,
prevalence of Alzheimer’s and other forms of dementia for example the number of neurons or synapses available
occurring in the senium, but evidence now available is to be lost differs among individuals. This idea was
sufficiently intriguing to warrant further study of the supported by study findings that suggested the prevalence
issue”. Kittner and colleagues2 suggested that adjustment or incidence of dementia was lower in individuals with
should be made for level of education when screening for larger brains than in those with smaller brains.13,14 I
dementia to avoid ascertainment bias, whereas Berkman3 suggest that this is a passive model of brain reserve—ie,
suggested that we must remain open to the view that a large brain might simply be able to tolerate more
“educational level and/or socioeconomic behavior pathology before it reaches the critical threshold for
correlated with it may be a genuine risk factor for senile clinical symptoms to appear. By contrast, cognitive
dementia and are worthy of scientific exploration in their reserve is an active form of reserve in which brain
own right”. Zhang and colleagues4 reported that a low function rather than brain size is the relevant variable.
level of education was associated with increased prevalence The concept of cognitive reserve suggests that the brain
of Alzheimer’s disease and dementia in a probability actively attempts to cope with pathology by using pre-
sample survey of 5055 older people not admitted to care existing cognitive-processing approaches or com-
facilities in Shanghai, China. These observations sparked pensatory mechanisms.15,16 Therefore, an individual with
my interest in studying the association between aspects of high cognitive reserve would cope better with the same
life experience and dementia; subsequently I have under- amount of pathology than an individual with low
taken a long-term research programme to investigate cognitive reserve, even when brain size is the same.
cognitive reserve. In this Personal View I present a Although the initial concept of brain reserve was
theoretical account of cognitive reserve, summarise entirely quantitative, several studies have suggested
epidemiological research that has lent support to the wider underlying biological features. For example,
concept, and describe imaging studies that have attempted stimulating environments have been associated with
to identify the neural substrates of cognitive reserve. I will neurogenesis17–19 and upregulation of BDNF, which
also discuss the potential clinical implications of the fosters neural plasticity and could impart reserve.20 Brain
concept of cognitive reserve. Although I discuss cognitive reserve and cognitive reserve, therefore, seem to make
reserve in the context of Alzheimer’s disease and normal independent and synergistic contributions to our
ageing, it has also been reported to provide benefit in understanding of individual differences in clinical
patients with vascular injury,5–7 Parkinson’s disease,8 resilience to brain pathology. Whether the two
traumatic brain injury,9 HIV,10 neuropsychiatric disorders,11 components of reserve interact remains unresolved.
and multiple sclerosis.12 Cognitive reserve was initially posited as a moderator
between brain change and clinical outcome, but life
Brain reserve and cognitive reserve experience may also prevent or minimise brain pathology.
The concept of reserve has been put forward to account On a simple level, physical exercise has long been
for differences between individuals in susceptibility to recognised to help prevent vascular disease. Likewise,
age-related brain changes and pathology, such as that participation in cognitively stimulating activities has
seen in Alzheimer’s disease. Reserve is purported to act been suggested to slow the rate of hippocampal atrophy
as a moderator between pathology and clinical outcome, in normal ageing,21 and perhaps even to prevent

1006 www.thelancet.com/neurology Vol 11 November 2012

 Personal View

accumulation of amyloid plaques.22 However, although rapid decline than those with low cognitive reserve.26,27
these ideas are promising and intriguing, they are We matched patients with Alzheimer’s disease according
beyond the scope of this Personal View. to clinical severity:26 those with greater educational and
occupational attainment died sooner than those with
Epidemiological evidence for cognitive reserve lower attainment. In a subsequent analysis, those with
My colleagues and I first investigated the concept of greater educational or occupational attainment had
cognitive reserve in a study of incident dementia, based more rapid cognitive decline.27 On average, scores on
on the assumption that Alzheimer’s disease pathology memory tests following diagnosis decline by about one
slowly develops over time independently of cognitive point every year in patients with low attainment, but by
reserve, and that the pathology begins to accumulate two points in those with higher attainment.27,28 A more
many years before the onset of clinically diagnosed rapid decline in cognitive function was also seen in
Alzheimer’s disease (figure 1).23 Because people with patients with Alzheimer’s disease who engaged in more
greater reserve should be able to tolerate more leisure activities before the onset of dementia than in
Alzheimer’s disease pathology, the onset of clinical those who engaged in few activities.29 Our theoretical
dementia in these individuals should be delayed. explanation for these findings is that individuals with
593 non-demented individuals aged 60 years or older high cognitive reserve can tolerate more pathology and,
were followed up for more than 4 years. Individuals with therefore, the point at which cognitive functions begin
less than 8 years of education had 2·2 times higher risk to be affected will be later than in those with lower
of developing dementia than those with more years of cognitive reserve. However, in all people a common
education. Occupational attainment was also assessed, point is reached when the pathology is so severe that
based on US census categories and classified as low function cannot be maintained (figure 1). According to
(unskilled, semiskilled, skilled trade or craft, and clerical these assumptions, individuals with the greatest
or office worker) or high (manager, business or cognitive reserve will have more advanced pathology at
government, and professional or technical) occupational the onset of cognitive decline, less time until they reach
levels. Participants with low lifetime occupational the point when pathology overwhelms function, and
attainment were at 2·25 times greater risk of developing thus a more rapid rate of decline.
dementia than those with high lifetime occupational Although these epidemiological studies support the
attainment. The implication of these findings was that concept of cognitive reserve, the evidence is not definitive.
educational and occupational experiences created a Only controlled studies can truly establish whether
reserve against the effects of Alzheimer’s disease interventions or experiences are beneficial.
pathology. Self-reported participation in various leisure
activities in the preceding month were assessed by
interview in a set of non-demented elderly participants.24
The effects of participation in 13 groups of activities High reserve
were assessed: knitting, listening to music, or other
hobby; walking for pleasure or excursion; visiting friends
or relatives; being visited by relatives or friends; physical
conditioning; going to movies, restaurants, or sporting
Cognitive test score

Low reserve
events; reading magazines, newspapers, or books;
watching television or listening to the radio; doing
unpaid community volunteer work; playing cards,
games, or bingo; going to a club or centre; going to
classes; and going to church, synagogue, or temple.
Score at incident AD visit
Participants were allocated according to low (six or fewer
activities) or high (more than six activities) participation
in leisure activities. Those who engaged in more than six
leisure activities had 38% lower risk of developing
dementia than participants who partook in fewer
activities. The authors of a review25 of 22 cohort studies Neuropathology
of the effects of education, occupation, premorbid IQ,
Figure 1: Hypothesised change in memory function over time in individuals with high and low cognitive
and mental activities in incident dementia found that
reserve
most of the studies reported a significant protective AD pathology begins to advance before changes in memory performance are observed. Decline is seen later in
effect of these lifetime exposures, and calculated that the individuals with high cognitive reserve because pathology is tolerated longer than by people with low cognitive
protective effect of higher cognitive reserve decreased reserve. The figure shows a hypothetical point at which pathology is so severe that memory performance is nil.
This point is the same for individuals with high and low reserve. The rate of decline, however, differs between
the risk of developing dementia by 46%. groups, and is more rapid in individuals with high reserve than in those with low reserve. According to this model,
By contrast, after clinical presentation of Alzheimer’s the differential rate of decline is seen irrespective of whether individuals have been diagnosed as having AD before
disease, patients with high cognitive reserve show more memory has begun to decline. AD=Alzheimer’s disease.

www.thelancet.com/neurology Vol 11 November 2012 1007

 Personal View

by 0·088 standard units. Thus, at any given level of brain

Severe AD

pathology, higher education was associated with better

cognitive function.

Neural mechanisms
The cognitive or neural mechanisms that might underlie
the reserve against age-related or Alzheimer’s disease-
High reserve related pathology are unknown. Functional imaging has
been used to try to identify networks that might mediate
Clinical severity

Low reserve cognitive reserve. On the basis of these studies, I propose

Mild AD

that neural implementation of cognitive reserve might

Diagnostic threshold
take two forms: neural reserve and neural compensation.16,38
Neural reserve is the idea that cognitive reserve could be
MCI

associated with differences between individuals in the

resilience of pre-existing cognitive networks, and neural
compensation is the idea that some individuals might be
Normal

better than others at using compensatory mechanisms.

Mild Moderate The key idea behind neural reserve is that cognitive
AD pathology reserve might be mediated by the same networks that are
used by individuals in the absence of pathology related to
Figure 2: Clinical implications of cognitive reserve in patients with AD
age or disease. For example, the differential efficiency or
Individuals with low cognitive reserve might seem to be clinically demented when AD pathology is mild, whereas
those with higher cognitive reserve might remain clinically normal. At higher levels of pathology, both groups capacity of these networks could account for variation
might appear to be clinically demented. Still, those with higher reserve will appear to be less clinically severe than between individuals in performance and the ability to
those with lower reserve. AD=Alzheimer’s disease. MCI=mild cognitive impairment. cope with brain changes. In a functional MRI (fMRI)
study, network efficiency and capacity were assessed in
Neuroimaging studies of cognitive reserve young (age 20–29 years) and old (age 60–69 years)
Resting regional cerebral blood flow adults.39 The extrinsic difficulty of a task was manipulated
Epidemiological studies suggest that at any given level of by shortening of the response deadline. Covariance
clinical severity in Alzheimer’s disease, the degree of analysis of the imaging data showed a spatial pattern of
pathology will be greater in individuals with higher activation in young and old adults that increased as the
cognitive reserve than in those with lower cognitive deadlines shortened. The spatial pattern was expressed
reserve (figure 2). This idea was tested by assessment of to a greater degree by the older adults when the deadline
resting regional cerebral blood flow as a surrogate for was longest. By contrast, the young adults’ pattern
Alzheimer’s disease pathology.30,31 In patients matched for expression was greater than the older adults’ at the
clinical severity, an inverse relation was found between shorter deadlines when the task was most difficult. This
resting regional cerebral blood flow and years of finding is consistent with the idea that network efficiency
education.32 Higher level of education was associated with and capacity are reduced with increasing age. Differences
greater depletion of blood flow in the parietotemporal in efficiency and capacity were also noted among
area (figure 3), where PET changes are seen in patients individuals within each age group; therefore, individuals
with Alzheimer’s disease. This observation provided an with more efficient or higher-capacity networks could
initial indication that patients with higher cognitive have more resilience in the face of age-related or disease-
reserve had more Alzheimer’s disease pathology than related changes.
those with lower cognitive reserve even though they Neural reserve and neural compensation were also
appeared clinically similar. Similar associations have assessed in a series of imaging studies.40,41 Young (mean
been shown for occupational attainment and leisure age 25·1 years) and older (74·4 years) healthy adults were
activities,33–36 and a post-mortem study confirmed the assessed with the letter Sternberg task. In this task,
relation with education.37 The investigators examined at participants study one, three, or six letters for 3 seconds
autopsy the brains of 130 elderly patients who had (stimulus phase) followed by a 7-second retention phase.
undergone cognitive assessment about 8 months before They are then presented with a single probe letter and are
death. Education and a summary measure of Alzheimer’s asked to indicate whether it was included in the previously
disease pathology (mean standardised density of neuritis studied set. fMRI analyses were used to assess load-related
and diffuse plaques and neurofibrillary tangles) were activation: the aspects of activation that change as the
associated with cognitive performance, and education number of letters studied increases. Multivariate linear
modified the association between Alzheimer’s disease modelling showed that load-related activation during the
pathology and levels of cognitive function: for each retention phase was characterised by two networks of
additional year of education, the relation between brain regions that seemed to work together as the task got
Alzheimer’s disease pathology and cognition was reduced harder,40 as inferred by an increase in fMRI signal. The

1008 www.thelancet.com/neurology Vol 11 November 2012

 Personal View

first network was used by both the young and elderly functioning, assessment of IQ or consideration of
adults and involved areas often associated with working occupation might be useful.
memory (eg, midline cerebellum and left insula/inferior Individuals with high cognitive reserve, by definition,
frontal gyrus). The second network was primarily will present with disease-related clinical symptoms later
characterised by activation in parahippocampal areas and than individuals with low cognitive reserve. This
was used consistently only in the elderly group. The difference could be identified in individuals with high
elderly adults who used the second network the most had cognitive reserve by the use of more challenging tests or
the worst performance in the task. In a follow-up analysis tests that are more specific for particular pathology (eg,
atrophy in a key area within the first network (the left associative learning tasks for the hippocampus).
precentral gyrus) was associated with decreased efficiency Clinicians must, however, be aware that even in
in that network and increased use of the second network.41 individuals with underlying pathology there will still be a
This observation might be an example of neural period during which sensitive measures cannot detect
compensation—ie, as age-related changes limit the cognitive changes in patients with high cognitive reserve.
efficiency of the first network, the second network Information on brain structure integrity could be
becomes increasingly used. Although those who rely integrated with cognitive data for diagnostic purposes.
more on the compensatory network can still perform the Neuroimaging tools have the potential to detect
task, they do so less successfully than those who rely on pathological changes when only slight impairment is
the first network. A further analysis42 identified two indicated by neuropsychological testing, particularly in
potential influences of cognitive reserve (as measured by individuals with high cognitive reserve who maintain
intelligence quotient [IQ]) in these individuals. First, the cognitive functioning for an extended period of time.
individuals with high cognitive reserve could tolerate For example, at a given level of clinical severity in
more atrophy in the first network and still preserve patients with Alzheimer’s disease, higher levels of
performance without having to resort to using the second
network. Second, individuals with high cognitive reserve
did better in the letter Sternberg task, even when they use
the second network, than those with low cognitive
reserve.42 This finding suggests that patients with high
cognitive reserve can make use of resources that are
Above HS AD
separate from those directly involved in task performance, 107
and is consistent with the idea that generalised neural 106
representation of cognitive reserve could impart protection 104
across a wide range of tasks. To investigate the possibility
103
of a generalised neural representation of cognitive reserve,
imaging data acquired during the stimulus-presentation 102
phase of two different tasks with different cognitive 101
demands were analysed. In young adults, an activation 99
network was identified, in which expression increased
98
with increasing load-related activation and correlated with
high cognitive reserve.43 This type of network could be an 97 HS graduate AD
example of the neural instantiation of cognitive reserve. In 96
the future, identification of neural patterns of activation 94
that are associated with generic cognitive reserve could
93
provide direct ways of measuring the level of reserve in
any individual. Furthermore, a quantifiable neural pattern 92
for cognitive reserve might lead to stratification of 91
outcomes in studies of pharmacological and non- 89
pharmacological approaches to improve cognitive
88
functioning.
Below HS AD
Application of cognitive reserve in clinical
assessment
When cognition is assessed as part of a diagnostic work-
up, the most appropriate validated indicators of cognitive
reserve for each patient—such as educational or Figure 3: Cerebral blood flow as a proxy for AD pathology in patients with different levels of education
Each group comprised 20 AD patients matched for clinical severity based on assessments of mental status and
occupational attainment—should be used. In the event activity of daily living scores. The lowest blood flow (dark blue) was seen in the parietotemporal areas of patients
that an individual’s level of education is not believed to with the highest educational level, which indicates the most advanced AD pathology. AD=Alzheimer’s disease.
be a good representation of his or her optimum cognitive HS=high school. Reproduced from Stern et al,32 by permission of John Wiley and Sons.

www.thelancet.com/neurology Vol 11 November 2012 1009

 Personal View

education are associated with more severe disease- assessed elderly people who played a complex role-playing
related changes on PET scans than seen for those with computer game for fifteen 1·5-h training sessions in the
lower levels of education.44,45 laboratory in a period of 4–5 weeks, resulting in a total
Integration of the cognitive reserve concept into the training time of 22·5 h. They reported improved
interpretation of biomarker imaging, however, is performance in a wide range of cognitive tasks. Similarly,
premature. Although the presence of amyloid in the focused computerised training improved working
brain can be detected with PET or testing of CSF, the memory capacity and extended working memory to tasks
prognostic implications of these tests are not fully in which participants received no training.51 The
established. These biomarkers could, however, provide limitations and promise of various cognitive interventions
fruitful avenues for research into cognitive reserve. in relation to lifelong learning and ageing have been set
Cognitive reserve should also be recognised as a factor out by a working group.52
that will affect the rate of cognitive decline after diagnosis. A complex computer game designed by human
The rate of decline after clinical onset is more rapid in operations psychologists has been used to test different
individuals with high reserve than in those with low approaches to training young adults in complex tasks.53
reserve, even when other factors that might contribute to One approach, called emphasis-change training, improved
the disease course are taken into account.27,28,46 This effect gameplay and enabled extension of training to other tasks.
has direct relevance for assessing the efficacy of potential Players were instructed to focus on all of the features of the
treatments in clinical trials, because the response to a game, but to give particular attention to one feature per
particular medication might be altered by the degree of game and to change to a different feature each time they
underlying pathology. Additionally, most clinical trials are played. For example, in some games they were instructed
designed to compare rates of decline between treatment to control the ship that they were piloting, but in others
and control groups. A mismatch in cognitive reserve they attended more to destroying mines that appeared
across groups could lead to differences in the rates of occasionally on the screen. By shifting emphasis from
decline that have nothing to do with the study drug. game to game, participants did not fall into a fixed strategy.
Finally, epidemiological evidence that links specific life Rather, they developed the ability to deal with the task as a
experiences and individual characteristics to reduced risk whole. This approach could be termed attention allocation
of dementia is insufficient to show definitively whether or executive control. By using this approach, young adults
these have any direct preventive or delaying effects. were able to use the training to incorporate a demand and
Intervention studies are needed to firmly establish causal simultaneously do an additional task, while playing the
links between these features and cognitive reserve. In the game.54 Ability in real-world training tasks (eg, performance
meantime, patients should be recommended to engage in flight simulator training) was also improved.55,56 In a
in mental and physical activities that are unlikely to be preliminary study the ability of older adults (mean age 66
harmful, and clinicians should be careful not to present years) to learn to play the game, and the possibility that
these activities as established treatments or proven gameplay might improve cognitive performance were
preventive strategies against dementia. investigated.57 90 individuals were assigned to three
groups: no playing, playing the game without emphasis-
Cognitive reserve in remediation and prevention change training, and playing the game with emphasis-
Epidemiological evidence suggests that experiences at change training. In the two game groups participants
all stages, even in later life, can contribute to cognitive played three times per week for 12 weeks. Although the
reserve. Intervention might, therefore, be useful even in game was challenging for the elderly participants, only six
elderly patients to impart or maintain reserve, slow age- dropped out of the study and gameplay performance
related cognitive decline, and prolong healthy ageing. improved with time in those who completed the study.
The most successful remediation approach so far has Players in the no-emphasis-change group were less
been aerobic exercise. Controlled studies have shown focused on the key goals of the game than those in the
that in elderly individuals with respiratory capacity emphasis-change group; they were more likely to respond
below the median at baseline, aerobic exercise increases to opportunities for bonus-point rewards than to consider
respiratory capacity and cognitive performance.47,48 strategy. By contrast, players in the emphasis-change
Cognitive intervention studies, however, have had group focused more on the central features of the game
mixed results. In one large study, participants received and on achieving the overall goal of attacking and
training in one of three cognitive domains: memory, destroying the fortress. The primary cognitive measures
reasoning, and speed of information processing. assessed were five tasks that involve executive control: trail-
Training in one cognitive domain did not affect making test, letter-number sequencing test, Stroop colour
outcomes in the two other domains. Additionally, and word test, set-switching task, and flanker task. On one
training did not result in any improvement in formal of these tasks, a test of working memory, emphasis-change
measures of everyday function.49 training was associated with greater improvement in
Studies that involve participants in complex gameplay working memory from baseline than in the other two
have shown promise. For example, Basak and colleagues50 training groups. On the basis of these findings, a study has

1010 www.thelancet.com/neurology Vol 11 November 2012

 Personal View

been started to assess the effects of playing complex games

in conjunction with aerobic exercise, with the hope of Search strategy and selection criteria
showing a synergistic effect through a boost in brain This Personal View focuses primarily on the work of the
reserve from exercise (eg, by improving plasticity via author and his group since 1992. It does not represent an
upregulation of BDNF) and a rise in cognitive reserve by exhaustive literature review on this topic. Other included
increasing the efficiency of the cognitive networks articles were already known to the author and are intended to
underlying executive control. support points made in the discussion.
How life experience contributes to the creation of
cognitive reserve and protection against age-related and
disease-related pathology remains unknown. Focused Again, a directly measurable neural correlate of reserve,
but large-scale studies are required to formulate specific such as an fMRI pattern, would be useful. More generally,
recommendations. The most meaningful endpoints consideration of cognitive and brain reserve will be
would be slowed rate of age-related decline in cognitive helpful in the development of systems-based approaches
function or reduced risk of developing Alzheimer’s to understanding brain changes. Since the brain passively
disease. No study has yet shown effects on either of these and actively attempts to cope with brain changes or
endpoints. The optimum cohorts would include elderly pathology, the factors that contribute to resilience need to
people with intact cognitive function followed up for be understood.
years. Studies could, and probably should, involve several Conflicts of interest
interventions, including exercise, cognitive stimulation, I declare that I have no conflicts of interest.
and social stimulation. A drawback of such studies is that Acknowledgments
they would be very expensive, will require a large number This work was funded by NIA RO1 AG14671.
of participants, and will have to be conducted over several References
years. In the meantime, a general recommendation is to 1 Gurland B. The borderlands of dementia: the influence of
sociocultural characteristics on rates of dementia occurring in the
continue educational and mentally stimulating activities senium. Aging 1981; 15: 61–84.
throughout life.58 2 Kittner SJ, White LR, Farmer ME, et al. Methodological issues in
screening for dementia: the problem of education adjustment.
J Chronic Dis 1986; 39: 163–70.
Conclusions 3 Berkman LF. The association between educational attainment and
The concept of cognitive reserve arose from epidemiological mental status examinations: of etiologic significance for senile
observations. Various life exposures seem to be associated dementias or not? J Chronic Dis 1986; 39: 171–74.
with resilience against age-related or pathology-related 4 Zhang M, Katzman R, Salmon D, et al. The prevalence of dementia
and Alzheimer’s disease in Shanghai, China: impact of age, gender
impairment of cognitive function. The original observations and education. Ann Neurol 1990; 27: 428–37.
indicated the involvement of easily measurable variables, 5 Elkins JS, Longstreth WT, Manolio TA, Newman AB, Bhadelia RA,
such as education or occupational attainment, but other Johnston SC. Education and the cognitive decline associated with
MRI-defined brain infarct. Neurology 2006; 67: 435–40.
lifestyle factors, for instance behaviours that stimulate
6 Dufouil C, Alperovitch A, Tzourio C. Influence of education on the
cognition, personality, and so on, also seem to be important. relationship between white matter lesions and cognition. Neurology
Overall contributions to reserve most likely come from 2003; 60: 831–36.
multiple sources, and cognitive reserve might, therefore, 7 Dufouil C, Alperovitch A, Ducros V, Tzourio C. Homocysteine,
white matter hyperintensities, and cognition in healthy elderly
change across an individual’s lifespan dependent on people. Ann Neurol 2003; 53: 214–21.
exposures and behaviours. This idea suggests that changes 8 Glatt SL, Hubble JP, Lyons K, et al. Risk factors for dementia in
in lifestyle and interventions might be useful to delay age- Parkinson’s disease: effect of education. Neuroepidemiology 1996;
15: 20–25.
related cognitive decline or dementia. Epidemiological 9 Kesler SR, Adams HF, Blasey CM, Bigler ED. Premorbid
studies, however, only describe correlations and do not test intellectual functioning, education, and brain size in traumatic
causation, and carefully controlled studies will be required brain injury: an investigation of the cognitive reserve hypothesis.
Appl Neuropsychol 2003; 10: 153–62.
to assess the effects of practical interventions. 10 Farinpour R, Miller EN, Satz P, et al. Psychosocial risk factors of
The desire to understand the neural basis of cognitive HIV morbidity and mortality: findings from the Multicenter AIDS
reserve has been a motivating factor behind functional Cohort Study (MACS). J Clin Exp Neuropsychol 2003; 25: 654–70.
imaging studies. The findings might contribute to the 11 Barnett JH, Salmond CH, Jones PB, Sahakian BJ. Cognitive reserve
in neuropsychiatry. Psychol Med 2006; 36: 1053–64.
understanding of age-related changes in brain behaviour 12 Sumowski JF, Chiaravalloti N, Deluca J. Cognitive reserve protects
and could lead to targeted interventions. Direct neural against cognitive dysfunction in multiple sclerosis.
measures of cognitive reserve, perhaps based on patterns J Clin Exp Neuropsychol 2009; 31: 913–26.
13 Katzman R, Robert T, DeTeresa R, et al. Clinical, pathological, and
derived from fMRI, would be very useful. neurochemical changes in dementia: a subgroup with preserved
The concept of cognitive reserve could be applied mental status and numerous neocortical plaques. Ann Neurol 1988;
during clinical assessment. The association between high 23: 138–44.
14 Schofield PW, Logroscino G, Andrews H, Albert S, Stern Y. An
cognitive reserve and rapid decline in patients with association between head circumference and Alzheimer’s disease in
Alzheimer’s disease makes it important to include this a population-based study of aging. Neurology 1997; 49: 30–37.
factor in diagnostic formulations and in clinical trials that 15 Stern Y. What is cognitive reserve? Theory and research application
of the reserve concept. J Int Neuropsychol Soc 2002; 8: 448–60.
compare rates of decline between drugs and control.

www.thelancet.com/neurology Vol 11 November 2012 1011

 Personal View

16 Stern Y. Cognitive reserve. Neuropsychologia 2009; 47: 2015–28. 38 Stern Y, Habeck C, Moeller J, et al. Brain networks associated with
17 Brown J, Cooper-Kuhn CM, Kemperman G, van Praag H, Winkler J, cognitive reserve in healthy young and old adults. Cereb Cortex
Gage FH. Enriched environment and physical activity stimulate 2005; 15: 394–402.
hippocampal but not olfactory bulb neurogenesis. Eur J Neurosci 39 Stern Y, Rakitin BC, Habeck C, et al. Task difficulty modulates
2003; 17: 2042–46. young-old differences in network expression. Brain Res 2012;
18 van Praag H, Christie BR, Sejnowski TJ, Gage FH. Running 1435: 130–45.
enhances neurogenesis, learning, and long-term potentiation in 40 Zarahn E, Rakitin B, Abela D, Flynn J, Stern Y. Age-related changes
mice. Proc Natl Acad Sci USA 1999; 96: 13427–31. in brain activation during a delayed item recognition task.
19 van Praag H, Kemperman G, Gage FH. Running increases cell Neurobiol Aging 2007; 28: 784–98.
proliferation and neurogenesis in the adult mouse dentate gyrus. 41 Steffener J, Brickman AM, Rakitin BC, Gazes Y, Stern Y. The impact
Nat Neurosci 1999; 2: 266–70. of age-related changes on working memory functional activity.
20 van Praag H, Kempermann G, Gage FH. Neural consequences of Brain Imaging Behav 2009; 3: 142–53.
environmental enrichment. Nat Rev Neurosci 2000; 1: 191–98. 42 Steffener J, Reuben A, Rakitin BC, Stern Y. Supporting performance
21 Valenzuela MJ, Sachdev P, Wen W, Chen X, Brodaty H. Lifespan in the face of age-related neural changes: testing mechanistic roles
mental activity predicts diminished rate of hippocampal atrophy. of cognitive reserve. Brain Imaging Behav 2011; 5: 212–21.
PLoS One 2008; 3: e2598. 43 Stern Y, Zarahn E, Habeck C, et al. A common neural network for
22 Landau SM, Marks SM, Mormino EC, et al. Association of lifetime cognitive reserve in verbal and object working memory in young
cognitive engagement and low β-amyloid deposition. Arch Neurol but not old. Cereb Cortex 2008; 18: 959–67.
2012; published online Jan 23. DOI:10.1001/archneurol.2011.2748. 44 Garibotto V, Borroni B, Kalbe E, et al. Education and occupation as
23 Stern Y, Gurland B, Tatemichi TK, Tang MX, Wilder D, Mayeux R. proxies for reserve in aMCI converters and AD: FDG-PET evidence.
Influence of education and occupation on the incidence of Neurology 2008; 71: 1342–49.
Alzheimer’s disease. JAMA 1994; 271: 1004–10. 45 Kemppainen NM, Aalto S, Karrasch M, et al. Cognitive reserve
24 Scarmeas N, Levy G, Tang MX, Manly J, Stern Y. Influence of hypothesis: Pittsburgh Compound B and fluorodeoxyglucose
leisure activity on the incidence of Alzheimer’s disease. Neurology positron emission tomography in relation to education in mild
2001; 57: 2236–42. Alzheimer’s disease. Ann Neurol 2008; 63: 112–18.
25 Valenzuela MJ, Sachdev P. Brain reserve and dementia: a systematic 46 Hall CB, Derby C, LeValley A, Katz MJ, Verghese J, Lipton RB.
review. Psychol Med 2005; 25: 1–14. Education delays accelerated decline on a memory test in persons
26 Stern Y, Tang MX, Denaro J, Mayeux R. Increased risk of mortality who develop dementia. Neurology 2007; 69: 1657–64.
in Alzheimer’s disease patients with more advanced educational 47 Angevaren M, Aufdemkampe G, Verhaar H, Aleman A, Vanhees L.
and occupational attainment. Ann Neurol 1995; 37: 590–95. Physical activity and enhanced fitness to improve cognitive function
27 Stern Y, Albert S, Tang MX, Tsai WY. Rate of memory decline in AD in older people without known cognitive impairment.
is related to education and occupation: cognitive reserve? Neurology Cochrane Database Syst Rev 2008; 2: CD005381.
1999; 53: 1942–57. 48 Kramer AF, Hahn S, Cohen NJ, et al. Ageing, fitness and
28 Scarmeas N, Albert SM, Manly JJ, Stern Y. Education and rates of neurocognitive function. Nature 1999; 400: 418–19.
cognitive decline in incident Alzheimer’s disease. 49 Jobe JB, Smith DM, Ball K, et al. ACTIVE: a cognitive intervention
J Neurol Neurosurg Psychiatry 2006; 77: 308–16. trial to promote independence in older adults. Control Clin Trials
29 Helzner EP, Scarmeas N, Cosentino S, Portet F, Stern Y. Leisure 2001; 22: 453–79.
activity and cognitive decline in incident Alzheimer disease. 50 Basak C, Boot WR, Voss MW, Kramer AF. Can training in a
Arch Neurol 2007; 64: 1749–54. real-time strategy video game attenuate cognitive decline in older
30 Friedland RP, Brun A, Budinger TF. Pathological and positron adults? Psychol Aging 2008; 23: 765–77.
emission tomographiccorrelations in Alzheimer’s disease. Lancet 51 Klingberg T. Training and plasticity of working memory.
1985; 325: 228. Trends Cogn Sci 2010; 14: 317–24.
31 McGeer EG, Peppard RP, McGeer PL, et al. ¹⁸Fluorodeoxyglucose 52 The Royal Society. Brain waves module 2: neuroscience:
positron emission tomography studies in presumed Alzheimer implications for education and lifelong learning. http://royalsociety.
cases, including 13 serial scans. Can J Neurol Sci 1990; 17: 1–11. org/uploadedFiles/Royal_Society/Policy_and_Influence/Module_2_
32 Stern Y, Alexander GE, Prohovnik I, Mayeux R. Inverse relationship Neuroscience_Education_Full_Report_Printer_Friendly.pdf
between education and parietotemporal perfusion deficit in (accessed Sept 9, 2012).
Alzheimer’s disease. Ann Neurol 1992; 32: 371–75. 53 Mane A, Donchin E. The space fortress game. Acta Psychol 1989;
33 Stern Y, Alexander GE, Prohovnik I, et al. Relationship between 71: 17–22.
lifetime occupation and parietal flow: implications for a reserve 54 Fabiani M, Buckley J, Gratton G, Coles MGH, Donchin E. The
against Alzheimer’s disease pathology. Neurology 1995; 45: 55–60. training of complex task performance. Acta Psychol 1989;
34 Scarmeas N, Zarahn E, Anderson KE, et al. Association of life 71: 259–99.
activities with cerebral blood flow in Alzheimer disease: 55 Gopher D, Weil M, Bareket T. Transfer of skill from a computer
implications for the cognitive reserve hypothesis. Arch Neurol 2003; game trainer to flight. Hum Factors 1994; 36: 387–405.
60: 359–65. 56 Gopher D, Weil M, Siegal D. Practice under changing priorities:
35 Alexander GE, Furey ML, Grady CL, et al. Association of premorbid an approach to the training of complex skills Acta Psychol 1989;
intellectual function with cerebral metabolism in Alzheimer’s 71: 147–77.
disease: implications for the cognitive reserve hypothesis. 57 Blumen HM, Gopher D, Steinerman JR, Stern Y. Training cognitive
Am J Psychiatry 1997; 154: 165–72. control in older adults with the space fortress game: the role of
36 Perneczky R, Drzezga A, Ehl-Schmid J, et al. Schooling mediates training instructions and basic motor ability. Front Aging Neurosci
brain reserve in Alzheimer’s disease: findings of 2010; 2: 145.
fluoro-deoxy-glucose-positron emission tomography. 58 Orrell M, Sahakian B. Education and dementia. BMJ 1995;
J Neurol Neurosurg Psychiatry 2006; 77: 1060–63. 310: 951–52.
37 Bennett DA, Wilson RS, Schneider JA, et al. Education modifies the
relation of AD pathology to level of cognitive function in older
persons. Neurology 2003; 60: 1909–15.

1012 www.thelancet.com/neurology Vol 11 November 2012