REVIEWS

The changing prevalence and

incidence of dementia over time
— current evidence
Yu‑Tzu Wu1, Alexa S. Beiser2, Monique M. B. Breteler3, Laura Fratiglioni4,
Catherine Helmer5, Hugh C. Hendrie6, Hiroyuki Honda7, M. Arfan Ikram8,
Kenneth M. Langa9, Antonio Lobo10, Fiona E. Matthews11, Tomoyuki Ohara12,
Karine Pérès5, Chengxuan Qiu4, Sudha Seshadri13, Britt-Marie Sjölund4, Ingmar Skoog14
and Carol Brayne15
Abstract | Dementia is an increasing focus for policymakers, civil organizations and
multidisciplinary researchers. The most recent descriptive epidemiological research into
dementia is enabling investigation into how the prevalence and incidence are changing over
time. To establish clear trends, such comparisons need to be founded on population-based
studies that use similar diagnostic and research methods consistently over time. This narrative
Review synthesizes the findings from 14 studies that investigated trends in dementia prevalence
(nine studies) and incidence (five studies) from Sweden, Spain, the UK, the Netherlands, France,
the USA, Japan and Nigeria. Besides the Japanese study, these studies indicate stable or declining
prevalence and incidence of dementia, and some provide evidence of sex-specific changes. No
single risk or protective factor has been identified that fully explains the observed trends, but
major societal changes and improvements in living conditions, education and healthcare might
have favourably influenced physical, mental and cognitive health throughout an individual’s
life course, and could be responsible for a reduced risk of dementia in later life. Analytical
epidemiological approaches combined with translational neuroscientific research could provide
a unique opportunity to explore the neuropathology that underlies changing occurrence of
dementia in the general population.

Correspondence to C.B.
Department of Public Health
and Primary Care, Cambridge
Institute of Public Health,
Forvie Site, University of
Cambridge, School of Clinical
Medicine, Cambridge
Biomedical Campus,
Cambridge CB2 0SR, UK.
carol.brayne@
medschl.cam.ac.uk
doi:10.1038/nrneurol.2017.63
Published online 12 May 2017;
corrected online 17 May 2017
Dementia has become an important public health, eco-
nomic, social and political issue, and attracts increasing
investment into research. According to estimates from
the World Alzheimer Report 2015, 46.8 million people
worldwide have dementia, and this number is expected
to increase to 74.7 million by 2030 and 131.5 million
by 2050 (REF. 1). In light of this predicted dementia
‘epidemic’ and consequent economic burden, the G8
dementia summit in 2013 and the WHO Ministerial
Conference in 2015 resulted in calls for global action
against dementia. The summit established a goal to
identify a cure or disease-­modifying therapy by 2025
(REFS 2,3). To date, most dementia research has focused
on the neurological features, pathophysiological mech-
anisms, and drug discovery. This basic science approach
has provided knowledge about dementia at the individ-
ual or biological level, but a predominantly reductionist
approach that focuses on single mechanisms does not
suffice to provide an understanding of the full spectrum
of dementia in the general population and to iden-
tify risk factors across different populations and life
courses4. These aspects can only be investigated fully
with population-based e­ pidemiological research.
Investigation of the changes in dementia incidence
and prevalence over time is challenging, as changes
in diagnostic criteria and other methodological vari­
ation can affect the prevalence and incidence esti-
mates. Primary evidence must, therefore, be founded
on population-­based studies that have consistent study
designs and measurement methods over time. In the
past 5 years, several population-based studies of demen-
tia have been published in which consistent research
methods were used, providing new insight into the
descriptive epidemiology of dementia and challen­ging
the accepted forecasts of increasing prevalence and
incidence.

NATURE REVIEWS | NEUROLOGY VOLUME 13 | JUNE 2017 | 327

©
2
0
1
7
M
a
c
m
i
l
l
a
n
P
u
b
l
i
s
h
e
r
s
L
i
m
i
t
e
d
,
p
a
r
t
o
f
S
p
r
i
n
g
e
r
N
a
t
u
r
e
.
A
l
l
r
i
g
h
t
s
r
e
s
e
r
v
e
d
.
REVIEWS

Key points study designs and methodologies, and classifying pri-

• Knowledge of changes in dementia occurrence can only be acquired through
population-based studies conducted at different time periods in representative
samples derived from the same populations
• Such studies must use diagnostic and research methods that are as similar as feasible
across time to enable valid comparisons
• We synthesize worldwide evidence from 14 such population-based studies across
Western Europe, the USA, Japan and Nigeria; most have reported declining or stable
prevalence and incidence with varying sex differences across countries
• No single risk or protective factor has explained these changes, but societal changes
in western societies have improved cognitive reserve and health status across the
lifecourse
• Integrating analytical epidemiological approaches and neuroscience within
population-based studies is key to understanding the changes observed, underlying
neurobiological mechanisms, and what policies might sustain such improvements
Findings from five population-based studies in
Western Europe have previously been reviewed5, but
in this Review, we synthesize worldwide evidence on
changes and trends in dementia prevalence and incidence
from up‑to‑date research. We first discuss the historical
context of previous studies of dementia prevalence and
incidence and consider in detail the challenges involved
in obtaining accurate estimates of changes across time.
We then bring together the findings of worldwide
population-­based studies, discussing variations in the
Author addresses
1
REACH: The Centre for Research in Ageing and Cognitive Health, Department of
Psychology, College of Life and Environmental Sciences, University of Exeter,
Washington Singer Building, Perry Road, Exeter EX4 4QG, UK.
2
Boston University School of Public Health, 72 East Concord St, B602 Boston,
Massachusetts 02118, USA.
3
German Center for Neurodegenerative diseases (DZNE), Population Health Sciences,
Sigmund-Freud-Straße 27, 53127 Bonn, Germany.
4
Aging Research Center, Department of Neurobiology, Care Sciences and Society,
Karolinska Institutet-Stockholm University, Gävlegatan 16, S-113 30, Stockholm, Sweden.
5
INSERM, ISPED, Centre INSERM, U1219 - Bordeaux Population Health Research Center,
Bordeaux, France.
6
Department of Psychiatry, Indiana University School of Medicine, 410 West 10th Street,
Indianapolis, Indiana 46202, USA.
7
Department of Neuropathology, Graduate School of Medical Sciences, Kyushu University,
3‑1‑1, Maidashi, Higashi‑ku, 812–8582, Fukuoka, Japan.
8
Department of Epidemiology, Erasmus University Medical Center, Wytemaweg 80,
3015 CN Rotterdam, Netherlands.
9
Department of Internal Medicine, Veterans Affairs Center for Clinical Management
Research, Institute for Social Research, and Institute for Healthcare Policy and Innovation,
University of Michigan, 2800 Plymouth Road, Ann Arbor, Michigan 48109–2800, USA.
10
Department of Psychiatry, Universidad de Zaragoza and Instituto de Investigación
Sanitaria Aragón, Zaragoza. CIBERSAM, Madrid, Spain.
11
Institute of Health and Society, Newcastle University, The Baddiley-Clark Building,
Richardson Road, Newcastle Upon Tyne, NE4 5PL, UK.
12
Department of Epidemiology and Public Health, Graduate School of Medical Sciences,
Kyushu University, 3‑1‑1, Maidashi, Higashi‑ku, 812–8582, Fukuoka, Japan.
13
Boston University School of Medicine, The Framingham Study, 72 East Concord Street,
B602 Boston, Massachusetts 02118, USA.
14
Centre for Ageing and Health AgeCap, Institute of Neuroscience and Physiology,
Sahlgrenska Academy, University of Gothenburg, Box 100, S-405 30, Gothenburg, Sweden.
15
Department of Public Health and Primary Care, Cambridge Institute of Public Health,
Forvie Site, University of Cambridge, School of Clinical Medicine, Cambridge Biomedical
Campus, Cambridge CB2 0SR, UK.
mary and secondary evidence on the basis of their
research ­methods. Finally, we consider possible reasons
for the observed changes in dementia prevalence and
incidence and how basic neuroscience and population-­
based studies must be combined to gain a more complete
­understanding of dementia and its evolution.
Investigation of dementia trends
Previous studies
Population-based studies of dementia epidemiology were
initiated in the 1980s to inform policy development on
population ageing and dementia6. These investigations
started by examining prevalence (the proportion of
people with dementia in a defined population) before
moving on, with longitudinal results, to incidence (the
occurrence of new cases over a specific period). Results
from studies carried out in Western Europe contributed
to the European Studies of Dementia (EURODEM)
reports7,8, which synthesized epidemiological measures
across European countries and had a substantial effect
on policy and research. This pan-European collabora-
tion was reconvened to combine research resources and
insight from old and new population-based cohorts to
update the descriptive epidemiology of dementia in con-
temporary older European populations9. In the USA,
several nationwide and regional studies of ageing cohorts
conducted since the 1960s have included measures of
cognitive function, although the diagnosis of dementia is
seldom included in these studies10. Nationwide estimates
of dementia prevalence in the USA have been based on
the extrapolation of results from various localities to the
total population in the USA11. In the late 1980s and early
1990s, a small number of epidemiological investigations
were also conducted beyond Western Europe and the
USA, in locations such as Australia, Canada, Japan, China
and Taiwan12. Before the turn of the century, data on
low-income and middle-income countries were lacking,
but many active studies are now being conducted in these
societies1. All of these worldwide studies have provided
population metrics of dementia that are widely used for
policy making and lobbying for awareness and resources.
Challenges
Descriptive population-based studies of dementia
have been conducted for over 30 years, but studies that
investigate changes in its prevalence or incidence over
time have emerged only within the past few years13,14.
Estimates of changes in incidence and prevalence have
been made with statistical modelling and systematic
reviews1,11,12, but the extent of comparable data on these
changes has been limited by inconsistent methodolo-
gies across studies15. Diagnostic criteria have changed
substantially during these decades. Different criteria
are known to identify different groups of patients with
dementia16, and any differences in the approach to diag-
nosis will affect estimates of prevalence and incidence.
Given that many studies of dementia prevalence and
incidence have used diverse diagnostic methods (as
well as different contexts and time points), comparabil-
ity even between geographical locations has been poor,

328 | JUNE 2017 | VOLUME 13 www.nature.com/nrneurol

©
2
0
1
7
M
a
c
m
i
l
l
a
n
P
u
b
l
i
s
h
e
r
s
L
i
m
i
t
e
d
,
p
a
r
t
o
f
S
p
r
i
n
g
e
r
N
a
t
u
r
e
.
A
l
l
r
i
g
h
t
s
r
e
s
e
r
v
e
d
.
 REVIEWS

and has been poorer still across time. These changes in
diagnostic boundaries combined with parallel increases
in awareness of dementia among the public and med-
ical professionals has led to earlier diagnosis on aver-
age17. Furthermore, a consensus diagnosis, even when
made with use of standardized data collection methods
and standardized diagnostic criteria, can be affected
by changes in a clinician’s perceptions of diagnostic
­thresholds and criteria over time18,19.
Ultimately, diagnosis of dementia remains heavily
influenced by clinical judgement, medical records and
characteristics of the individuals that make up a study
population. For this reason, dementia remains a clinical
syndrome, with an emphasis on cognitive and functional
states, and identifying subtypes is particularly challenging.
For example, Alzheimer-type dementia can be diagnosed
with varying levels of investigation, but regardless of the
investigations used, diagnosis is based on the assump-
tion that the underlying pathology is Alzheimer-type
pathology. Subtype analysis is, therefore, more difficult
to keep consistent over time than is syndromic diagno-
sis, limiting the possibility of making valid comparisons
of prevalence or incidence between dementia subtypes.
Primary evidence
We identified 14 population-based studies in which the
study methods used at all time points were sufficiently
similar and in which the geographical areas were suffi-
ciently well-defined to enable a meaningful comparison.
Of these 14 studies, nine investigated prevalence and five
investigated incidence. The populations included were
from the USA, Western Europe, Japan and Nigeria.
Five other studies have also investigated trends in
dementia prevalence or incidence but were not included
because of methodological aspects that mean they can-
not be meaningfully compared with the other 14 studies
(BOX 1). Studies that focused on only Alzheimer dis-
ease20,21 or cognitive impairment22,23 were also excluded,
Box 1 | Excluded studies of dementia prevalence and incidence
Five studies have investigated trends in dementia prevalence and incidence but used
methods that prevent them from being meaningfully compared with others;
consequently, these studies were not included in our Review. Two Swedish studies81,82
were excluded. In one, prevalence and incidence trends during 1947–1957 and
1957–1972 were investigated81, and findings from these periods are likely to have
limited relevance to contemporary older populations. The other focused on short-term
prevalence trends in populations aged ≥85 years82, but had methodological flaws: the
study cohorts were not sampled independently, the analysis did not account for the
overlapping of the study population, and medical records were used to support
dementia diagnosis, which might have led to bias owing to differences in the
diagnostic boundaries applied for patients across time.
One Japanese study70 that investigated prevalence trends between 1980 and 2000
was excluded because the screening for dementia was based on self-reported cognitive
problems rather than objective cognitive testing, and clinical diagnosis was only
applied to those who reported cognitive problems. This approach might have led to
biased prevalence estimates because a marked change in awareness of dementia
within the population is likely to have affected patient self-reporting. Another report,
from the Hisayama study in Japan69, was excluded because it focused on an autopsy
subsample without clear representation of the older population in the study area.
One study has compared prevalence and incidence trends in dementia in China, but
used different diagnostic criteria at the two time points studies83, so was also excluded.
owing to heterogeneous definitions and diagnostic
methods across time and studies. Those using medical
records, healthcare administrative databases, systematic
reviews and meta-analyses are described briefly in the
latter part of this review.
Trends in prevalence
Study details. Nine studies have investigated trends in
dementia prevalence in France (the Bordeaux farmer
study)19, Sweden (the Gothenburg, Nordanstig and
Stockholm studies)24–26, Spain (the Zaragoza study)27,
the UK (the Cognitive Function and Ageing Study
(CFAS))28, the USA (the Health and Retirement Study
(HRS) and the Indianapolis–Ibadan Dementia Project
(IIDP))29,30, and Japan (the Hisayama study)31 (TABLE 1).
The earliest of these is the Gothenburg study (1976–
1977)24 and the most recent is the HRS (conducted in
2012)30. The Zaragoza study, the CFAS and the IIDP had
similar designs: two independent cohorts were recruited
across two time points in defined geographical areas27–29.
In the Gothenburg study24, age-specific prevalence at
ages 70 years and 75 years were compared over three
decades in random samples of local populations. In the
Nordanstig25 and Stockholm26 studies, the prevalence of
dementia from the earlier investigations (the Nordanstig
Project (1995–1998) and the Kungsholmen Project
(1987–1989)) was compared with the regional preva-
lence from a more recent nationwide cohort (Sweden
National Study on Aging and Care (SNAC; 2001–2004)).
The Bordeaux farmer study included only farmers living
in the Bordeaux area19. The HRS is a dynamic study in
which new cohorts are enrolled every 6 years to ensure
a representative sample of older adults in the USA30.
The only study from East Asia, the Hisayama study con-
ducted in Japan31, included all residents of the study area
who were aged ≥65 years at four time points.
In the Zaragoza study, the CFAS, the IIDP and the
Bordeaux farmer study, considerable drops in response
rate were seen across the two cohorts27–29. A wide range
of sensitivity analyses was conducted in the CFAS and
the Bordeaux farmer study to address potential selection
bias owing to this differential response rate, and revealed
limited effects on the results. In the Zaragoza study, a sam-
pling strategy was used to account for the population that
did not respond, so the estimates are considered to be rep-
resentative of the entire older population in the study area.
The nine studies had either one-stage or two-stage
designs: a one-stage design which included only a diag-
nosis phase that involved detailed examination and
application of clinical criteria, and a two-stage design
included a screening phase to identify potential par-
ticipants with dementia as well as a diagnostic phase.
Clinical diag­noses were mainly based on the Diagnostic
and Statistical Manual of Mental Disorders, third edition
revised (DSM‑III‑R) criteria32. Algorithmic diagnosis
was used in the CFAS, and algorithmic historical crite-
ria were used in the Gothenburg study; these approaches
were similar to the DSM‑III‑R criteria. In the French
study, an algorithm approach based on Mini-Mental
State Examination (MMSE)33 and Instrument Activity
of Daily Living (IADL)34 scores was used in addition

NATURE REVIEWS | NEUROLOGY VOLUME 13 | JUNE 2017 | 329

©
2
0
1
7
M
a
c
m
i
l
l
a
n
P
u
b
l
i
s
h
e
r
s
L
i
m
i
t
e
d
,
p
a
r
t
o
f
S
p
r
i
n
g
e
r
N
a
t
u
r
e
.
A
l
l
r
i
g
h
t
s
r
e
s
e
r
v
e
d
.
REVIEWS

to clinical diagnosis to identify potential participants
with dementia. Diagnosis in the HRS was based on a
27‑point cognitive test, or a proxy assessment if the
participant was unable to complete the interview30. The
assessment was conducted as a phone or face‑to‑face
interview, and was validated in a subsample of the HRS
cohort (the Aging, Demographics, and Memory Study
(ADAMS)); this approach had a 78% concordance with
clinical diagnosis made by medical professionals on the
basis of c­ riteria developed for the study35.

Table 1 | Designs and methodologies of the dementia prevalence studies

Study Study Cohorts Study Diagnostic methods Major changes Notes on changes
population designs*
Gothenburg24, People aged • 1: 1976–1977 (n = 707, R = 79%) One-stage Clinical diagnosis Subjective clinical N/A
Sweden 70 years and • 2: 2000–2001 (n = 579, R = 66%) (Historical criteria, opinion (changes
75 years in • 3: 2005–2006 (n = 753, R = 63%) similar to DSM-III‑R) in clinicians’
Gothenburg perception
of diagnostic
threshold)
Nordanstig25, People aged • 1: 1995–1998 (n = 303, R = 90%) One-stage Clinical diagnosis Subjective clinical Same physicians
Sweden ≥78 years in • 2: 2001–2003 (n = 384, R = 77%) (DSM-III‑R) opinion conducted diagnosis
Nordanstig to reduce subjective
clinical opinions
Stockholm26, People aged • 1: 1987–1989 (n = 1,700, • Two-stage Clinical diagnosis • Study designs Same physicians
Sweden ≥75 years in R = 72%) (cohort 1) (DSM-III‑R) • Subjective conducted diagnosis
Kungsholmen, • 2: 2001–2004 (n = 1,575, • One-stage clinical opinions to reduce subjective
Stockholm R = 73%) (cohort 2) clinical opinions
Zaragoza27, People aged • 1: 1987–1989 (n = 1,080, Two-stage Clinical diagnosis • Response rate Suggested that
Spain ≥65 years in R = 95%) (DSM-III‑R) • Subjective refusals might not
Zaragoza • 2: 1994–1996 (n = 3,715, clinical opinion affect the results (the
R‡ = 64%) sampling strategy
accounted for
non-response)
CFAS28, UK People aged • 1: 1991–1994 (n = 7,635, • Two-stage Algorithmic diagnosis • Study design • One-stage
≥65 years R = 80%) (cohort 1) (GMS-AGECAT, • Response rate interview validated
in England • 2: 2008–2011 (n = 7,796, • One-stage similar to DSM-III‑R) in the cohort 1
(Newcastle, R = 56%) (cohort 2) follow‑up
Nottingham, • Sensitivity analysis
Cambridgeshire) used to address
low response rate
in cohort 2
Bordeaux Farmers aged • 1: 1988–1989 (n = 595, R = 69%) Two-stage • Clinical diagnosis • Subjective • Same physicians
farmer19, ≥65 years in • 2: 2007–2008 (n = 906, R = 52%) (DSM-III‑R) clinical opinions conducted
France Bordeaux • Algorithmic • Response rate consensus
diagnosis diagnosis
(MMSE + IADL) • Sensitivity analysis
used to address
low response rate
HRS30, USA People aged • 1: 2000 (n = 10,546, R = 88%) One-stage Algorithmic Study design Increased
≥65 years in the • 2: 2012 (n = 10,516, R = 89%) diagnosis (phone face-to-face
USA, nationwide or face-to-face interview in cohort 2
interview, and reduced phone
27‑item cognitive interview and proxy
test or proxy assessment
assessment + IADL)
IIDP29, USA African- • 1: 1992 (n = 1,500, R‡ = 86%) Two-stage Clinical diagnosis • Difference Same basic group of
American people • 2: 2001 (n = 1,892, R = 44%) (DSM-III‑R, ICD‑10) in diagnostic clinicians from both
aged ≥70 years threshold across sites conducted
in Indianapolis, clinicians consensus process
USA • Response rate
Hisayama31, All residents • 1: 1985 (n = 2,457, R = 95%) Two-stage Clinical diagnosis • Screening and Changes in
Japan aged ≥65 years • 2: 1992 (n = 1,189, R = 97%) (DSM-III/DSM-III‑R) diagnostic methodologies
in Hisayama • 3: 1998 (n = 1,437, R = 100%) methods validated in cohorts
• 4: 2005 (n = 1,566, R = 92%) • Subjective
clinical opinion
*One-stage design involves a diagnosis phase only, two-stage design involves a screening phase and a diagnosis phase. ‡Response rate from the original cohorts
including younger age groups (Zaragoza, age ≥60 years; IIDP, age ≥65 years). CFAS, Cognitive Function and Ageing Study; DSM, Diagnostic and Statistical Manual;
DSM‑III‑R, DSM-III revised; GMS-AGECAT, Geriatric Mental State – Automated Geriatric Examination for Computer Assisted Taxonomy; HRS, Health and
Retirement Study; IADL, Instrument Activity of Daily Living; ICD‑10, International Statistical Classification of Diseases and Related Health Problems 10th Revision;
IIDP, Indianapolis–Idaban Dementia Project; N/A, not applicable; R, response rate.

330 | JUNE 2017 | VOLUME 13 www.nature.com/nrneurol

©
2
0
1
7
M
a
c
m
i
l
l
a
n
P
u
b
l
i
s
h
e
r
s
L
i
m
i
t
e
d
,
p
a
r
t
o
f
S
p
r
i
n
g
e
r
N
a
t
u
r
e
.
A
l
l
r
i
g
h
t
s
r
e
s
e
r
v
e
d
.
 REVIEWS

The intention in each of these studies was to imple-
ment the same diagnostic methods over time, but
changes in subjective clinical opinion cannot be ruled
out as a major factor that might have influenced case
identification and prevalence estimates. Measures were
taken in several of the studies to address this issue. In
the Nordanstig and Stockholm studies, the same physi-
cians made the diagnoses in the two cohorts within each
study. In the IIDP, a clinical consensus process was used,
in which the same group of clinicians made diagnoses
in the two cohorts. In the CFAS and the Gothenburg
study, algorithmic diagnosis with a structured psychi-
atric interview was used to avoid variation in subjective
opinions. Small changes in study designs and method-
ologies were made in the Hisayama study31, Stockholm
study26 and the CFAS28; to ensure that these changes had
minimal effects on prevalence estimates, the new meas-
urements used in these studies were tested and validated
before their application to subsequent cohorts.
Findings. In contrast to the increase in prevalence of
dementia that is predicted on the basis of projection
methods, most of the nine population-based studies
reported stable or declining prevalence over time (FIG. 1;
age-specific and sex-specific prevalence estimates are
in Supplementary information S1 (table)). The three
Swedish studies indicate generally stable prevalence
within the total population, albeit with wide confidence
intervals in all but the Stockholm study. The CFAS
demonstrates a 23% reduction in prevalence over two
decades in the total study population in England28, and
the HRS suggests a 26% decrease in prevalence over
12 years in the older US population30. The Bordeaux
farmer study found conflicting results when the two
diagnostic approaches were used: with the algorithmic
diagnosis, a 40% decline in prevalence was seen, but when
clinical diagnosis was used, an increase in prevalence
by more than twofold was seen19. The Nordanstig and
Zaragoza studies did not indicate significant reductions
in prevalence in the total population, but demonstrated
decreases of >50% among men25,27. Taken together, and
given the fact that the longevity of people with dementia
is increasing26, these results suggest an actual decline in
the age-specific risk of dementia.
Trends in incidence
Study details. Five studies have investigated trends in the
incidence of dementia in the Netherlands (the Rotterdam
study)36, France (the Bordeaux study)37, the UK (the
CFAS)38, the USA (the IIDP and the Framingham Heart
Study (FHS))39,40, and Nigeria (the IIDP)39 (TABLE 2). The
IIDP examined trends in African American people in
Indianapolis, USA, and in a Yoruba population in Ibadan,
Nigeria39. The Bordeaux incidence study focused on
urban residents rather than farmers, but used the same
reference cohort as the prevalence study37. The CFAS
and IIDP reported both prevalence and incidence trends
within the same study cohorts28,29,38,39.
In the Bordeaux study, the CFAS, and the IIDP,
incidence was measured in two independent cohorts,
whereas the Rotterdam study36 involved analysis of
non-overlapping sub-cohorts, and the FHS40 involved
analysis of participants from two generations. In the
Rotterdam study, the 1990 cohort included all residents
of the study area aged 60–90 years, and the 2000 cohort
included a non-overlapping sample of residents who had
since entered that age range or were of the correct age
and had moved into the study area. The FHS combined
data from an original cohort and a cohort of their off-
spring, and divided them into four epochs to compare
incidence across these periods. The follow‑up periods
and intervals varied between studies. To address differ-
ential response rate and potential effect of missing data,
several s­ ensitivity models were tested in the Bordeaux
study and CFAS.
Different sets of diagnostic criteria were used in
the Rotterdam study (DSM-III‑R), IIDP (DSM-III‑R,
ICD‑10) and FHS (DSM‑IV)32,41,42. In the Bordeaux
study and the CFAS, algorithmic diagnosis was used.
In the Bordeaux study37, MMSE33 and IADL34 scores
were used to define dementia, and in the CFAS38, a dif-
ferential diagnosis procedure derived from a structured
psychiatric interview was used. The Bordeaux study also
included clinical diagnosis, but different clinical criteria
were applied to the two cohorts, so these data were not
used to assess temporal trends.
Findings. Despite different study designs and meth-
ods, all five studies suggest a decrease in the incidence
of dementia in the total population across cohorts and
time periods (FIG. 2; age-specific and sex-specific inci-
dence estimates are in Supplementary information S2,S3
(tables)), although notable differences were seen in
subpopulations, particularly between the sexes. In the
Bordeaux study, the decrease was mainly driven by an
effect in women, whereas in the CFAS, the reduction
was confined to men. In the FHS, the reduction occurred
earlier and was sustained across the three epochs in
women, but occurred only in the last epoch in men.
The results of the IIDP suggest that the incidence was
reduced in African American people over a 10‑year
period, and indicate a 20% reduction in incidence in
Nigerian cohorts, but this effect was not statistically
significant. In the Bordeaux study, the results obtained
with clinical diagnosis differed from those obtained with
algorithm d ­ iagnosis; only the latter resulted in a decrease
in incidence37.
Secondary evidence
Secondary evidence for trends in dementia prevalence
and incidence is provided by studies that are based on
medical records, healthcare and insurance administra-
tive databases, systematic reviews and meta-analyses,
as these types of research do not provide the option of
controlling for changes in diagnostic methods, subjective
clinical opinions and public awareness. Several studies
conducted in Western Europe and North America
have reported trends in the prevalence or incidence of
dementia on the basis of analysis of medical records or
healthcare administrative databases22,43–52. These studies
tend to include large populations and are often based on
patients’ contact with medical services over time. These

NATURE REVIEWS | NEUROLOGY VOLUME 13 | JUNE 2017 | 331

©
2
0
1
7
M
a
c
m
i
l
l
a
n
P
u
b
l
i
s
h
e
r
s
L
i
m
i
t
e
d
,
p
a
r
t
o
f
S
p
r
i
n
g
e
r
N
a
t
u
r
e
.
A
l
l
r
i
g
h
t
s
r
e
s
e
r
v
e
d
.
REVIEWS

Cohort Population Odds ratio/prevalence

ratio (95% CI)

Gothenburg (age 70 years)

2000 vs 1976 Total 1.2 (0.5, 3.0)
2000 vs 1976 Men 0.5 (0.1, 3.1)
2000 vs 1976 Women 1.7 (0.6, 5.1)

Gothenburg (age 75 years)

2005 vs 1976 Total 1.2 (0.7, 2.2)
2005 vs 1976 Men 1.0 (0.4, 2.3)
2005 vs 1976 Women 1.4 (0.6, 3.3)

Nordanstig (age ≥78 years)

2001 vs 1995 Total 0.7 (0.5, 1.0)
2001 vs 1995 Men 0.5 (0.2, 0.9)
2001 vs 1995 Women 0.9 (0.5, 1.4)

Stockholm (age ≥75 years)

2001 vs 1987 Total 1.2 (0.9, 1.5)
2001 vs 1987 Men 0.9 (0.6, 1.5)
2001 vs 1987 Women 1.2 (1.0, 1.6)

Zaragoza (age ≥65 years)

1994 vs 1987 Total 0.8 (0.6, 1.0)
1994 vs 1987 Men 0.4 (0.3, 0.6)
1994 vs 1987 Women 1.0 (0.7, 1.5)

CFAS (age ≥65 years)

2008 vs 1991 Total 0.7 (0.6, 0.9)
2008 vs 1991 Men 0.6 (0.5, 0.8)
2008 vs 1991 Women 0.8 (0.7, 0.9)

Bordeaux farmer (age ≥65 years,

clinical diagnosis)
2007 vs 1988 Total 2.3 (1.5, 3.4)

Bordeaux farmer (age ≥65 years,

algorithmic diagnosis)
2007 vs 1988 Total 0.6 (0.4, 0.7)

HRS (age ≥65 years)

2012 vs 2000 Total 0.8 (0.7, 0.9)
2012 vs 2000 Men 0.7 (0.6, 0.8)
2012 vs 2000 Women 0.8 (0.7, 0.9)

IIDP (age ≥70 years)

2001 vs 1992 Total 1.1 (0.9, 1.4)

Hisayama (age ≥65 years)

1992 vs 1985 Total 0.7 (0.5, 1.0)
1998 vs 1985 Total 0.8 (0.6, 1.1)
2005 vs 1985 Total 1.3 (1.0, 1.9)
1992 vs 1985 Men 0.7 (0.5, 1.2)
1998 vs 1985 Men 0.8 (0.5, 1.3)
2005 vs 1985 Men 1.4 (0.9, 2.1)
1992 vs 1985 Women 0.6 (0.3, 1.3)
1998 vs 1985 Women 0.7 (0.4, 1.3)
2005 vs 1985 Women 1.3 (0.7, 2.2)

0.25 0.5 1.0 2.0 4.0

Figure 1 | Odds ratios and prevalence ratios reported in nine studies of dementia prevalence. Nature Reviews
Reported | Neurology
figures are the
ratios of the prevalence estimate for new cohorts to those for old cohorts. If prevalence estimates remain the same across two
cohorts, the ratio is 1.0. If estimates are higher in the new cohort than the old cohort, the ratio is greater than 1.0. Estimates
are adjusted as follows: Gothenburg study and HRS, unadjusted; IIDP, adjusted for age; Nordanstig, Zaragoza, Bordeaux and
Hisayama studies, adjusted for age and sex; Stockholm study, adjusted for age, sex and education; CFAS, adjusted for age,
sex, area and deprivation. In the Bordeaux study, clinical diagnosis was made by neuropsychologists using criteria in the
Diagnostic and Statistical Manual III revised, and algorithmic diagnosis was based on cognitive and functional ability tests.
CFAS, Cognitive Function and Ageing Study; HRS, Health and Retirement Study; IIDP, Indianapolis–Ibadan Dementia Project.

332 | JUNE 2017 | VOLUME 13 www.nature.com/nrneurol

©
2
0
1
7
M
a
c
m
i
l
l
a
n
P
u
b
l
i
s
h
e
r
s
L
i
m
i
t
e
d
,
p
a
r
t
o
f
S
p
r
i
n
g
e
r
N
a
t
u
r
e
.
A
l
l
r
i
g
h
t
s
r
e
s
e
r
v
e
d
.
 REVIEWS

analyses have mainly focused on short-term trends over
fewer than 10 years, and advanced analytical strategies
have been required to estimate prevalence or incidence
over continuous time periods and overlapping study
populations. Ascertainment bias, owing to inclusion
of patients who approached medical services, and dif-
ferences in diagnostic practice between clinical settings
(such as different departments and centres) cannot be
fully addressed in these analyses, making interpretation
of the findings challenging. Some of these studies have
suggested stability or reductions in the annual preva-
lence or incidence of dementia22,43–47, whereas others have
reported increases in prevalence or incidence47–52.
Owing to a lack of comparable data, trends in the
prevalence of dementia outside of western countries have
been estimated mainly on the basis of systematic reviews
and meta-analyses, which aggregate estimates from indi-
vidual studies conducted around the same time period.
Systematic reviews of numerous prevalence studies con-
ducted in East Asian countries have suggested that the
prevalence of dementia is increasing in Japan53, South
Korea54, Hong Kong55, Taiwan56 and China57,58, although
the increase in China is no longer significant when
methodological factors, including changes in diagnostic
criteria, are controlled for1,59,60. Two recent studies have
used different analytical methods to investigate changes
in cognitive impairment (measured by the MMSE) in the
Chinese Longitudinal Healthy Longevity Surveys, a large
cohort across 22 provinces in China. One study reports
declining incidence between 1998 and 2014 in people
aged ≥60 years61, whereas the other suggests increasing
prevalence across the oldest old cohorts aged ≥80 years62.
These differences might be related to selection of study
populations and variation in analytical approaches.
Whether any changes reported in East Asia are attribut-
able to heterogeneity of study designs and implementa-
tion or differences between high-income and low-income
countries, or eastern and western society remains unclear.
The different results obtained from primary investiga-
tions and systematic reviews emphasize the substantial
impact that changes in diagnostic methods, study designs
and social contexts over time can have on estimates of
epidemiological measurements.
Current evidence on dementia trends
Diagnosis of dementia, as of any disorder, is contextual
and can change across time and geographies. The second-
ary evidence discussed above was based on the analysis of
healthcare administrative databases, medical records, sys-
tematic reviews and meta-analyses, and the mixed find-
ings emphasize the fact that comparisons should not rely
on an overview of reported numbers, but must include
careful appraisal of methodologies and study contexts.
Changes in diagnostic methods, medical knowledge and
public awareness all influence identification of who meets
and does not meet study diagnostic criteria for dementia.

Table 2 | Designs and methodologies of the dementia incidence studies

Study Study Cohorts Study Diagnostic Major changes Response to changes
population design methods
Rotterdam36, All residents aged • 1: 1990 (n = 5,727, R = 73%) • Two-stage Clinical diagnosis Subjective N/A
the 60–90 years in • 2: 2000 (n = 1,769, R = 67%) • 3–4‑year (DSM-III‑R) clinical opinion
Netherlands Ommoord district follow‑up
of Rotterdam until 2007
Bordeaux37, People aged • 1: 1988–1989 (n = 1,469, • Two-stage • Clinical • Diagnostic • Algorithmic diagnosis
France ≥65 years in urban R = 60%) • Follow‑up diagnosis criteria used to compare with
Bordeaux • 2: 1999–2000 (n = 2,104, every (DSM-III‑R/-IV) • Response rate clinical diagnosis
R = 39%) 2–3 years • Algorithmic • Sensitivity models
for 10 years diagnosis conducted to address
(MMSE + IADL) differential response rates
CFAS38, UK People aged • 1: 1991–1994 (n = 7,635, • Two-stage Algorithmic • Study design • Imputation used to
≥65 years R = 80%) (cohort 1) diagnosis • Response rate address study design
in England • 2: 2008–2011 (n = 7,796, • One-stage (GMS-AGECAT, issue in cohort 1
(Newcastle, R = 56%) (cohort 2) similar to • Sensitivity models
Nottingham, • 2‑year DSM-III‑R) conducted to test effect
Cambridgeshire) follow‑up of missing data
IIDP39, USA African-American Indianapolis • Two-stage Clinical diagnosis • A clinical Same basic group of
and Nigeria people aged • 1: 1992 (n = 1,440, R‡ = 86%) • Follow‑up (DSM-III‑R, consensus clinicians from both sites
≥70 years in • 2: 2001 (n = 1,835, R = 44%) every ICD‑10) process conducted the consensus
Indianapolis, 2–3 years involving process
Ibadan
USA, and Yoruba until 2009 clinicians from
• 1: 1992 (n = 1,174, R‡ = 98%)
aged ≥70 years in both sites
• 2: 2001 (n = 1,895, R = 100%)
Ibadan, Nigeria • Response rate
FHS40, USA Longitudinal • Epoch 1: 1977–1983 (n = 2,457) • Two-stage Clinical diagnosis Subjective N/A
cohorts of people • Epoch 2: 1986–1991 (n = 2,135) • 5‑year (DSM‑IV) clinical opinion
aged ≥60 years in • Epoch 3: 1992–1998 (n = 2,333) follow‑up
Framingham • Epoch 4: 2004–2008 (n = 2,090)
*One-stage design involves a diagnosis phase only, two-stage design involves a screening phase and a diagnosis phase. ‡Response rate from the original cohorts
including younger age groups (age ≥65 years). CFAS, Cognitive Function and Ageing Study; DSM, Diagnostic and Statistical Manual; DSM‑III‑R, DSM-III revised;
FHS, Framingham Heart Study; GMS-AGECAT, Geriatric Mental State – Automated Geriatric Examination for Computer Assisted Taxonomy; IADL, Instrument
Activity of Daily Living; ICD‑10, International Statistical Classification of Diseases and Related Health Problems 10th Revision; IIDP, Indianapolis–Idaban Dementia
Project; MMSE, Mini Mental State Examination; N/A, not applicable; R, response rate.

NATURE REVIEWS | NEUROLOGY VOLUME 13 | JUNE 2017 | 333

©
2
0
1
7
M
a
c
m
i
l
l
a
n
P
u
b
l
i
s
h
e
r
s
L
i
m
i
t
e
d
,
p
a
r
t
o
f
S
p
r
i
n
g
e
r
N
a
t
u
r
e
.
A
l
l
r
i
g
h
t
s
r
e
s
e
r
v
e
d
.
REVIEWS

Cohort Population Hazard ratio/incidence

ratio (95% CI)
Rotterdam (age 60-90 years)
2000 vs 1990 Total 0.8 (0.6, 1.0)
2000 vs 1990 Men 0.7 (0.4, 1.2)
2000 vs 1990 Women 0.8 (0.5, 1.1)

Bordeaux
(clinical diagnosis, age ≥65 years)
1999 vs 1988 Total 0.9 (0.7, 1.1)
1999 vs 1988 Men 1.2 (0.8, 1.9)
1999 vs 1988 Women 0.9 (0.7, 1.2)

Bordeaux
(algorithmic diagnosis, age ≥65 years)
1999 vs 1988 Total 0.6 (0.5, 0.8)
1999 vs 1988 Men 1.1 (0.7, 1.8)
1999 vs 1988 Women 0.6 (0.5, 0.8)

CFAS (age ≥65 years)

2008 vs 1991 Total 0.8 (0.6, 1.0)
2008 vs 1991 Men 0.6 (0.4, 0.9)
2008 vs 1991 Women 1.0 (0.7, 1.3)

IIDP
(African American age ≥70 years)
2001 vs 1992 Total 0.4 (0.3, 0.5)

IIDP (Yoruba age ≥70 years)

2001 vs 1992 Total 0.8 (0.6, 1.1)

Framingham Heart Study

(age ≥60 years)
1986–1991 vs 1977–1983 Total 0.8 (0.6, 1.0)
1992–1998 vs 1977–1983 Total 0.6 (0.5, 0.8)
2004–2008 vs 1977–1983 Total 0.6 (0.4, 0.8)
1986–1991 vs 1977–1983 Men 1.0 (0.6, 1.6)
1992–1998 vs 1977–1983 Men 0.9 (0.6, 1.4)
2004–2008 vs 1977–1983 Men 0.6 (0.4, 1.1)
1986–1991 vs 1977–1983 Women 0.7 (0.5, 1.0)
1992–1998 vs 1977–1983 Women 0.5 (0.4, 0.7)
2004–2008 vs 1977–1983 Women 0.5 (0.4, 0.8)

0.25 0.5 1.0 2.0 4.0

Figure 2 | Hazard ratio and incidence rate ratio from five studies of dementia incidence. Reported figures are the ratios of
the incidence estimate in new cohorts to that in old cohorts. If incidence estimates remain the same across two cohorts, the
Nature Reviews
ratio is 1.0. If estimates are higher in new cohorts than old cohorts, the ratio is greater than 1.0. Estimates | Neurology
are adjusted as
follows: Rotterdam study, IIDP and Bordeaux study adjusted for age; Framingham Heart Study adjusted for age and sex; CFAS
adjusted for age, sex, area and deprivation. In the Bordeaux study, clinical diagnosis was made by neuropsychologists and
neurologists using criteria from the Diagnostic and Statistical Manual III revised and V, and algorithmic diagnosis was based on
cognitive and functional ability tests. CFAS, Cognitive Function and Ageing Study; IIDP, Indianapolis–Ibadan Dementia Project.

The different results obtained with use of clinical and
algorithmic diagnoses in the two Bordeaux studies19,37 dis-
cussed above further emphasize the effects of changes in
diagnostic boundaries and the substantial effect they can
have on estimates of prevalence and incidence over time.
Identification of true prevalence and incidence
trends, therefore, requires population-based studies that
use similar research methods across different time peri-
ods; the 14 primary studies discussed above are all such
studies. In many of these studies, declines in response
rates and changes in diagnostic boundaries used to make
consensus diagnoses were reported, factors that are
likely to affect the results. Nevertheless, despite differ-
ent study designs, methodologies and settings between
the individual studies, these population-based studies
generally report stabilization of, or a reduction in, the
prevalence and incidence of dementia. Given that mor-
tality is declining in the general population, stabiliza-
tion of dementia prevalence also suggests a decline in
the incidence. This primary evidence indicates a reduc-
tion in dementia occurrence in more recent generations,
­particularly in western countries.
Possible explanations
Western countries
The primary evidence from western countries discussed
in this Review suggests a reduction in the risk of devel-
oping dementia and improved health in old age across

334 | JUNE 2017 | VOLUME 13 www.nature.com/nrneurol

©
2
0
1
7
M
a
c
m
i
l
l
a
n
P
u
b
l
i
s
h
e
r
s
L
i
m
i
t
e
d
,
p
a
r
t
o
f
S
p
r
i
n
g
e
r
N
a
t
u
r
e
.
A
l
l
r
i
g
h
t
s
r
e
s
e
r
v
e
d
.
 REVIEWS

Box 2 | Factors related to decreasing dementia occurrence
The four studies listed here investigated the effects of risk and protective factors on
decrease in dementia prevalence or incidence. The common factors included in the
analysis for these studies were education, smoking, hypertension, cardiovascular disease
(stroke or heart disease), diabetes mellitus, BMI and cholesterol levels.
Bordeaux study, France26
• Adjusted for factors to test whether the observed decreases in dementia prevalence
and incidence were attenuated.
• Education and vascular factors had a small effect, but the trends of decreasing
prevalence and incidence remained significant.
Framingham Heart Study, USA40
• Adjusted for factors to test whether the observed decrease in dementia incidence
was attenuated.
• No significant effect of any investigated factor — changes were <10%
Health and Retirement Study, USA30
• Adjusted for factors to test whether the observed decrease in dementia prevalence
was attenuated.
• Education, cardiovascular factors and BMI attenuated the change by up to 12%, but
the decrease in prevalence remained significant.
Rotterdam Study, Netherlands64
• Calculated population attributable risk (PAR; the proportion of dementia cases that
could be prevented if the risk factor was removed) for each risk factor in the two
cohorts studied.
• The PAR was reduced by smoking and cholesterol levels.
• The PAR was unaffected by education and cardiovascular disease.
• The PAR was increased by diabetes mellitus and hypertension.
generations. Indeed, in the Rotterdam study, improve-
ment of brain health — as indicated by larger brain
volume, less brain atrophy and less cerebral small ves-
sel disease — was reported in the most recent cohort36.
Several possible explanations for these findings have
been suggested5,63, but only four of the studies30,37,40,64
identified the key factors associated with the trend of
decreasing dementia incidence (BOX 2), and these factors
vary between American, French and Dutch cohorts.
Educational level explained some of the decline in
incidence: up to 6% in the FHS and nearly 10% in the
Bordeaux incidence study. In the HRS, education along
with other socioeconomic factors explained 10% of the
decrease in prevalence. However, in the Rotterdam study,
the percentage of preventable dementia cases related to
low education remained similar over two decades, sug-
gesting that, despite improvements in education in the
more recent cohort, the level of education still contrib-
uted substantially to dementia occurrence in the more
recent cohort64.
The change in the proportion of cases that were pre-
ventable by cessation of smoking partly explained the
changes in incidence in the Rotterdam study, but smok-
ing did not explain the decline in incidence in the FHS
and Bordeaux study.
In these four studies, the changing prevalence of sev-
eral chronic diseases that are associated with dementia
— such as stroke, heart diseases, hypertension and dia-
betes mellitus — was reported but differed across stud-
ies. These changes explain only a limited proportion
(<10%) of the reduction in incidence and prevalence of
dementia, although any positive effects would be subject
to time lags, the lengths of which are unknown. Most
medications for cardiovascular disease and other chronic
diseases (­antihypertensives that act on systolic blood pres-
sure, anti­depressants and statins) have only been widely
available since the 1990s, so their effect on the observed
changes in dementia prevalence and incidence might
be limited; further improvements in these treatments
for chronic conditions might change the risk of devel-
oping dementia in later life and might, therefore, have
further effects on dementia incidence and prevalence in
the future46,65. Other lifestyle factors, such as changes in
diet and physical activity, have been suggested as reasons
for a declining incidence of dementia, but primary evi-
dence to confirm such hypotheses is currently lacking,
and these patterns are changing with each generation.
Some of the prevalence and incidence studies
revealed sex differences in the changes (TABLE 3), and
these differences could provide further insight into pos-
sible reasons. Life expectancy at the age of 60 years is a
good marker of the overall health status of elderly people
(based on UN data), and despite the fact that women
in western countries have consistently had a longer life
expectancy at the age of 60 years, the life expectancy of
men has increased to a greater degree over the past two
decades. Reductions in smoking and improvements
in the prevention and treatment of cardiovascular dis-
ease might have had a larger effect on the health and
life expectancy of men than those of women, and such
major changes in risk factors might be important in the
observed sex difference in dementia occurrence.
Although the reasons for stable or decreasing preva-
lence and incidence of dementia remain unclear, a sin-
gle risk factor is unlikely to be responsible. Changes in
Western societies and improvements in living conditions
since the two World Wars have led to improvements in
overall health and in cognitive development and reserve
throughout the lifecourse66. Population-level invest-
ments in infrastructure, education, health services and
social welfare have contributed to substantial improve-
ments in multiple dimensions of physical, mental and
cognitive health from early life and might have conse-
quently reduced the risk of dementia in later life. For
example, education level has been related to an increase
in cognitive reserve67. A higher level of education has
been associated with better cognitive performance, but
not with lower rates of cognitive decline68. The latest gen-
erations to reach old age have had more years of statu-
tory education than previous generations, which might
be associated with greater cognitive reserve, which, in
turn, might partly explain a delayed onset of dementia.
Such effects on incidence can only be observed over
decades. Nevertheless, the observations we have indi-
cate that factors that are related to social disadvantage
and health inequality might have an important role in
cognitive health over the lifecourse.
Beyond Western countries
Primary evidence that provides information about the
prevalence and incidence of dementia outside Western
Europe and the USA is lacking. Systematic reviews and

NATURE REVIEWS | NEUROLOGY VOLUME 13 | JUNE 2017 | 335

©
2
0
1
7
M
a
c
m
i
l
l
a
n
P
u
b
l
i
s
h
e
r
s
L
i
m
i
t
e
d
,
p
a
r
t
o
f
S
p
r
i
n
g
e
r
N
a
t
u
r
e
.
A
l
l
r
i
g
h
t
s
r
e
s
e
r
v
e
d
.
REVIEWS

Table 3 | Studies in which sex differences were reported for trends in the prevalence and incidence of dementia
Study Prevalence trends Incidence trends Life expectancy at age 60 years‡ (years)
Men Women Men Women Men Women
1990 2000 2012 1990 2000 2012
CFAS, UK (2008 vs 1991) ↓* ↓* ↓* ↔ 18 20 22 22 23 25
Zaragoza, Spain (1994 vs 1987) ↓* ↔ N/A N/A 19 21 22 24 25 27
Nordanstig, Sweden (2001 vs 1995) ↓* ↔ N/A N/A 19 21 23 23 24 25
Bordeaux, France (1999 vs 1988) N/A N/A ↔ ↓* 20 20 23 25 26 27
FHS, USA (2005 vs 1985) N/A N/A ↓ ↓* 19 20 21 23 23 24
↓ Decrease. ↔ Stable. *Trend was statistically significant. ‡Based on WHO data. CFAS, Cognitive Function and Ageing Study; FHS, Framingham Heart Study;
N/A, not applicable.

meta-analyses can be used to synthesize evidence on
dementia prevalence and incidence in low and mid-
dle income countries, but such analyses of trends are
unlikely to be robust if variations in methodologies and
population characteristics are not taken into account5.
Of the primary, population-based studies discussed
in the previous sections, two were conducted outside
Western Europe and the USA: one in Japan and one in
Nigeria, and these reported different trends.
The Hisayama study reported an increasing preva-
lence of dementia between 1985 and 2005, and findings
from a subsample of autopsy specimens suggested an
even higher prevalence in 2012 (REFS 31,69). However,
the analysis of the autopsy subsample did not take into
account age and potential selection bias. Another study
conducted in the Diasen-Cho area of Japan, which
was excluded from the discussion in previous sections
because the screening approaches used did not ensure
representation of the whole population, also identified
an increase in prevalence across three time points (1980,
1990 and 2000)70. By contrast, the incidence of dementia
in the Nigerian cohort of the IIDP was stable39.
A unifying explanation for these different results
is difficult to formulate because these countries have
different economic, political, social and cultural back-
grounds and rates of change over the past few dec-
ades. Some insight might be gained by considering the
changes in life expectancy at birth over the past century
(FIG. 3). Changes in life expectancy are associated with
the effects of societal factors and might therefore also
indicate different determinants of cognitive health across
generations71. Different life events, health statuses and
disease profiles experienced by different generations,
and trends in the prevalence and incidence of dementia
might reflect complex interactions between these fac-
tors. The life expectancy profiles of Japan and Nigeria
over the past century differ dramatically from those
of the western countries. In Japan, the life expectancy
was lower than that in western countries in the first
half of the 20th century and increased dramatically in
the 1960s. Life expectancy in Nigeria has increased by
30 years over the past century, but is still 20 years lower
than the other countries. These dramatic changes are
generally related to war and historical events, with a con-
sequence of extremely deprived living conditions, inter-
ruption of education and lack of health and social care in
the early life of the study cohorts. The long-term effects
of these societal factors can lead to variation in popu-
lation health, cognitive reserve and dementia occur-
rence across generations in countries. Future trends in
dementia occurrence outside western countries are less
predictable because the interplay between health during
the lifecourse, protective factors and risk factors varies
considerably between social contexts.
The Hisayama study includes longitudinal data on
cardiovascular disease in middle-aged cohorts, which might
also provide some insight into potential reasons for the
observed changes in dementia prevalence in Japan. Since
the mid‑1980s, the prevalence of hypertension, stroke and
smoking has declined, and the prevalence of diabetes mel-
litus, hypercholesterolaemia and obesity has increased72.
The increasing prevalence of dementia might, therefore,
be related to changes in lifestyle, such as an increasingly
western diet, physical inactivity and increasing obe-
sity, metabolic syndromes and diabetes mellitus31,73. An
analysis published in 1995 showed that these factors
were not associated with an increased risk of all-type
dementia after a 7‑year follow‑up, and a subtype analysis
found only one significant association, between diabetes
mellitus and vascular dementia, the prevalence of which
appeared to be stable31,74. However, a more recent analysis
after a 15‑year follow‑up showed that diabetes mellitus
was related to an increased risk of all-type dementia and
Alzheimer disease, but not with vascular dementia75.
Although these findings indicate a long-term effect of
diabetes mellitus on dementia in old age and a possible
time-lagged effect on increasing prevalence, increasing
recognition of mixed dementia will make interpretation
of these changes in the occurrence of dementia subtypes
over time even more challenging. Until deeper pheno-
typing, both during life and after death, is conducted and
is consistent across time, the detail of neurobiological
changes that underlie the risks and clinical manifestations
of dementia will be unknown.
Population ageing and an increasing burden of non-
communicable diseases are important challenges world-
wide, but the effect of chronic diseases on dementia
trends might vary between contexts, with time-lags of
uncertain length. Forecasts of dementia burden must take
into account these different contexts of health profiles,
deprivation and social environments rather than only
focusing on the potential effect of individual risk factors.

336 | JUNE 2017 | VOLUME 13 www.nature.com/nrneurol

©
2
0
1
7
M
a
c
m
i
l
l
a
n
P
u
b
l
i
s
h
e
r
s
L
i
m
i
t
e
d
,
p
a
r
t
o
f
S
p
r
i
n
g
e
r
N
a
t
u
r
e
.
A
l
l
r
i
g
h
t
s
r
e
s
e
r
v
e
d
.
 REVIEWS

100
World War I Great Depression World War II African American Civil End of Cold
(1914–1918) (1929–1937) (1939–1945) Rights Movement (1960s) War (1991)
90
Famines in Continental Europe

80 Flu pandemic Polio epidemic

(1918–1920) (1949)

70
Life expectancy (age)

60

50

40

30
Meiji restoration Nuclear bombs Japanese economic Japanese asset
(1868–1912) in Japan (1945) miracle (1960s) bubble collapse (1991)
20
End of British HIV/AIDS
Nigeria (1960) (1990s)
10 Unification of Niger areas
under British Colony Nigeria Civil War Nigeria democratization
(1914) (1967–1970) (1999)
0
1900 1905 1910 1915 1920 1925 1930 1935 1940 1945 1950 1955 1960 1965 1970 1975 1980 1985 1990 1995 2000 2005 2010 2015
Year

Spain Sweden Netherlands United France Japan Nigeria United United States:
Kingdom States black

Figure 3 | Life expectancy at birth in all countries included in population-based studies of dementia incidence and
Nature
prevalence. Data obtained from from Gapminder, the National Center for Health Statistics, USA 84
andReviews Neurology
Wu, Y.‑T.| et al.5.

Neuroscience and epidemiology symptoms and needs to be accompanied by deep pheno-

Basic neuroscience research into the pathology and
treatment of dementia has largely remained separate
from epidemiological research into dementia. However,
both areas of research have limitations, and taking an
approach that combines them will probably improve our
overall understanding of dementia.
Current basic research into dementia largely focuses
on mechanistic aspects to inform the development
of treatment, potential biomarkers for the diagnosis of
dementia subtypes before clinical signs develop, and
treatment efficacy in highly selected clinical samples76.
However, population-based studies have repeatedly
shown serious inconsistencies in the association between
the degree of neuropathology and cognitive performance,
as well as considerable overlap of pathology in people
with and without dementia77,78, demonstrating that our
understanding of the pathology that underlies dementia
remains incomplete. New techniques for defining brain
pathology and so‑called normal function might, there-
fore, be needed, and these techniques must be grounded
in research within contemporary populations so as to
provide insight into the neurobiology that underlies the
population changes in dementia incidence and preva-
lence that population-based studies have demonstrated.
Observational risk factor analysis can only go so far
towards the identification of factors related to dementia
typing that can be mapped back to populations in order
to provide further understanding of underlying neuro­
biological mechanisms. An illustration of this need is that
population-based studies seem to have identified sex-
related differences in the changes in dementia occur-
rence, but basic research will be needed to understand the
neuroscience that underlies these differences. In addition,
data on different populations, such as migrants, aborig-
inal populations and disadvantaged sectors of society,
are clearly lacking from the existing population-based
­studies79,80. New cohorts will be needed to provide a
comprehensive picture of brain health across global
­populations and to address health inequalities.
The concept of population-based studies — that is,
recruiting participants from community-based con-
texts to ensure representations of the whole population
— should be incorporated into future neurobiological
and neuropathological research in dementia. Results
from small, clinic-based samples, which include only
patients from memory clinics or other medical services,
have inherently limited generalizability and considerable
potential for bias owing to highly selective recruitment.
In particular, people who are socially disadvantaged are
less likely to take part in such research. Given o
­ ngoing
changes in brain health in the general population, the
integration of neuroscience with population-based

NATURE REVIEWS | NEUROLOGY VOLUME 13 | JUNE 2017 | 337

©
2
0
1
7
M
a
c
m
i
l
l
a
n
P
u
b
l
i
s
h
e
r
s
L
i
m
i
t
e
d
,
p
a
r
t
o
f
S
p
r
i
n
g
e
r
N
a
t
u
r
e
.
A
l
l
r
i
g
h
t
s
r
e
s
e
r
v
e
d
.
REVIEWS

studies and analytical epidemiological approaches
(neuroscientific epidemiological approaches) is vitally
important and provides an opportunity to integrate
understanding of brain health, neurobiology and neuro­
pathology within the general population to support
better prevention, care and cure of dementia.
Conclusions
In the past few years, descriptive epidemiological stud-
ies, in which the effects of changing diagnostic criteria
and study methods were minimized, have strengthened
the evidence that dementia — age for age — is declining
in some countries and that the number of people with
dementia can remain stable despite population age-
ing28,30. Substantial reductions in the risk of dementia in
whole populations can balance the growing numbers of
older people. Given that the global population is ageing,
identifying the factors that contribute to reductions in
the prevalence and incidence of dementia in particu-
lar countries and regions should become a major pri-
ority, as the findings will have important implications
on health and social policies in relation to dementia
prevention and risk reduction.
No single factor has been identified that fully explains
the observed changes in dementia prevalence and inci-
dence, but reductions in absolute inequalities — including
improvements in living conditions, better access to educa-
tion and improved healthcare systems — are likely to have
influenced multiple risk and protective factors throughout
an individual’s lifecourse that are related to physical, men-
tal and cognitive health, and thereby reduced the risk of
dementia in later life. This conclusion sends an important
message to all in society about the need for long-term
action to address factors that determine both healthy
and unhealthy ageing, and to make further efforts to
reduce inequalities within and between nations, with the
expectation of health with age, including a lower risk of
dementia. Only an integrated approach that incorporates
lifecourse health and brings together many disciplines
underpinned by neuroscience and population-based
epidemiological studies can provide the robust evidence
required to understand the observed changes.

1. Prince, M. et al. World Alzheimer Report 2015. The
global impact of dementia: an analysis of prevalence,
incidence, cost and trends. Alzheimer’s Disease
International https://www.alz.co.uk/research/
WorldAlzheimerReport2015.pdf (2015).
2. Department of Health. G8 dementia summit
declaration. GOV.UK https://www.gov.uk/
government/publications/g8-dementia-summit-
agreements/g8-dementia-summit-declaration
(2013).
3. World Health Organization. First WHO ministerial
conference on global action against dementia.
World Health Organization http://www.who.int/
mental_health/neurology/dementia/ministerial_
conference_2015_report/en/ (2015).
4. Brayne, C. & Davis, D. Making Alzheimer’s and
dementia research fit for populations. Lancet 380,
1441–1443 (2012).
5. Wu, Y.‑T. et al. Dementia in Western Europe:
epidemiological evidence and implications for policy
making. Lancet Neurol. 15, 116–124 (2016).
6. Brayne, C., Stephan, B. C. M. & Matthews, F. E.
European perspective on population studies of
dementia. Alzheimers Dement. 7, 3–9 (2011).
7. Hofman, A. et al. The prevalence of dementia in
Europe: a collaborative study of 1980–1990 findings.
Int. J. Epidemiol. 20, 736–748 (1991).
8. Fratiglioni, L. et al. Incidence of dementia and major
subtypes in Europe: a collaborative study of
population-based cohorts. Neurologic Diseases in the
Elderly Research Group. Neurology 54, S10–S15
(2000).
9. EU Joint Programme – Neurodegenerative Disease
Research. 21st century EURODEM. Redefining
dementia for epidemiology. JPND research http://
www.neurodegenerationresearch.eu/wp-content/
uploads/2015/02/21st-Century-EURODEM.pdf
(2014).
10. Bell, J. F. et al. Existing data sets to support studies of
dementia or significant cognitive impairment and
comorbid chronic conditions. Alzheimers Dement. 11,
622–638 (2015).
11. Alzheimer Association. 2015 Alzheimer’s disease facts
and figures. Alzheimers Dement. 11, 332–384 (2015).
12. World Health Organisation & Alzheimer’s Disease
International. Dementia: a public health priority.
World Health Organization http://www.who.int/
mental_health/publications/dementia_report_2012/en/
(2012).
13. Sachdev, P. S. Is the incidence of dementia declining?
A report for Alzheimer’s Australia. Alzheimer’s
Australia https://www.fightdementia.org.au/sites/
default/files/Paper_39_Is_the_incidence_of_dementia_
declining.pdf (2014).
14. Prince, M. et al. Recent global trends in the prevalence
and incidence of dementia, and survival with
dementia. Alzheimers Res. Ther. 8, 23 (2016).
15. Roehr, S., Pabst, A., Luck, T. & Riedel-Heller, S.G.
Secular trends in the incidence of dementia in high-
income countries: a protocol of a systematic review and
a planned meta-analysis. BMJ Open 7, e013630 (2017).
16. Erkinjuntti, T., Østbye, T., Steenhuis, R. & Hachinski, V.
The effect of different diagnostic criteria on the
prevalence of dementia. N. Engl. J. Med. 337,
1667–1674 (1997).
17. Grimmer, T. et al. Trends of patient referral to a memory
clinic and towards earlier diagnosis from 1985–2009.
Int. Psychogeriatr. 27, 1939–1944 (2015).
18. Kukull, W. A. et al. Interrater reliability of Alzheimer’s
disease diagnosis. Neurology 40, 257–260 (1990).
19. Pérès, K. et al. Trends in the prevalence of dementia
in French farmers from two epidemiological cohorts.
J. Am. Geriatr. Soc. 65, 415–420 (2017).
20. Hebert, L. et al. Change in risk of Alzheimer disease
over time. Neurology 75, 786–789 (2010).
21. Rocca, W. A. et al. Trends in the incidence and
prevalence of Alzheimer’s disease, dementia, and
cognitive impairment in the United States. Alzheimers
Dement. 7, 80–93 (2011).
22. Manton, K., Gu, X. & Ukraintseva, S. Declining
prevalence of dementia in the U.S. elderly population.
Adv. Gerontol. 16, 30–37 (2005).
23. Langa, K. M. et al. Trends in the prevalence and
mortality of cognitive impairment in the United States:
is there evidence of a compression of cognitive
morbidity? Alzheimers Dement. 4, 134–144 (2008).
24. Wiberg, P., Waern, M., Billstedt, E., Östling, S. &
Skoog, I. Secular trends in the prevalence of dementia
and depression in Swedish septuagenarians
1976–2006. Psychol. Med. 43, 2627–2634 (2013).
25. Wimo, A. et al. Cohort effects in the prevalence and
survival of people with dementia in a rural area in
Northern Sweden. J. Alzheimers Dis. 50, 387–396
(2016).
26. Qiu, C., von Strauss, E., Bäckman, L., Winblad, B.
& Fratiglioni, L. Twenty-year changes in dementia
occurrence suggest decreasing incidence in central
Stockholm, Sweden. Neurology 80, 1888–1894
(2013).
27. Lobo, A. et al. Prevalence of dementia in a southern
European population in two different time periods:
the ZARADEMP Project. Acta Psychiatr. Scand. 116,
299–307 (2007).
28. Matthews, F. E. et al. A two-decade comparison of
prevalence of dementia in individuals aged 65 years
and older from three geographical areas of England:
results of the Cognitive Function and Ageing Study I
and II. Lancet 382, 1405–1412 (2013).
29. Hall, K. S. et al. Prevalence rates for dementia and
Alzheimer’s disease in African Americans: 1992 versus
200. Alzheimers Dement. 5, 227–233 (2009).
30. Langa, K. et al. A comparison of the prevalence of
dementia in the United States in 2000 and 2012.
JAMA Intern. Med. 177, 51–58 (2017).
31. Sekita, A. et al. Trends in prevalence of Alzheimer’s
disease and vascular dementia in a Japanese
community: the Hisayama Study. Acta Psychiatr.
Scand. 122, 319–325 (2010).
32. American Psychiatric Association. Diagnostic and
Statistical Manual of Mental Disorders 3rd edn
revised (American Psychiatric Association, 1987).
33. Folstein, M., Folstein, S. & McHugh, P. R. A practical
method for grading the cognitive state of patients for
the clinician. J. Psychiatr. Res. 12, 189–198 (1975).
34. Lawton, M. P. & Brody, E. M. Assessment of older
people: self-maintaining and instrumental activities
of daily living. Gerontologist 9, 179–186 (1969).
35. Crimmins, E. M., Kim, J. K., Langa, K. M. & Weir, D. R.
Assessment of cognition using surveys and
neuropsychological assessment: the Health and
Retirement Study and the Aging, Demographics, and
Memory Study. J. Gerontol. B Psychol. Sci. Soc. Sci.
66, i162–i171 (2011).
36. Schrijvers, E. M. C. et al. Is dementia incidence
declining? Trends in dementia incidence since 1990 in
the Rotterdam Study. Neurology 78, 1456–1463
(2012).
37. Grasset, L. et al. Trends in dementia incidence:
evolution over a 10‑year period in France. Alzheimers
Dement. 12, 272–280 (2016).
38. Matthews, F. E. et al. A two decade dementia incidence
comparison from the Cognitive Function and Ageing
Studies I and II. Nat. Commun. 7, 11398 (2016).
39. Gao, S. et al. Dementia incidence declined in African-
Americans but not in Yoruba. Alzheimers Dement. 12,
244–251 (2016).
40. Satizabal, C. L. et al. Incidence of dementia over
three decades in the Framingham Heart Study.
N. Engl. J. Med. 374, 523–532 (2016).
41. World Health Organization. The ICD‑10 classification
of mental and behavioural disorders: clinical
descriptions and diagnostic guidelines. World Health
Organization http://www.who.int/classifications/icd/en/
bluebook.pdf (1992).
42. American Psychiatric Association. Diagnostic and
Statistical Manual of Mental Disorders 4th edn
(American Psychiatric Association, 2000).
43. Doblhammer, G., Fink, A. & Fritze, T. Short-term
trends in dementia prevalence in Germany between
the years 2007 and 2009. Alzheimers Dement. 11,
291–299 (2015).
44. Doblhammer, G., Fink, A., Zylla, S. & Willekens, F.
Compression or expansion of dementia in Germany?
An observational study of short-term trends in
incidence and death rates of dementia between
2006/07 and 2009/10 based on German health
insurance data. Alzheimers Res. Ther. 7, 66 (2015).
45. Rocca, W. A., Cha, R. H., Waring, S. C. & Kokmen, E.
Incidence of dementia and Alzheimer’s disease:
a reanalysis of data from Rochester, Minnesota,
1975–1984. Am. J. Epidemiol. 148, 51–62 (1998).

338 | JUNE 2017 | VOLUME 13 www.nature.com/nrneurol

©
2
0
1
7
M
a
c
m
i
l
l
a
n
P
u
b
l
i
s
h
e
r
s
L
i
m
i
t
e
d
,
p
a
r
t
o
f
S
p
r
i
n
g
e
r
N
a
t
u
r
e
.
A
l
l
r
i
g
h
t
s
r
e
s
e
r
v
e
d
.
 REVIEWS

46. Sposato, L. A. et al. Declining incidence of stroke and
dementia: coincidence or prevention opportunity?
JAMA Neurol. 72, 1529–1531 (2015).
47. Kosteniuk, J. G. et al. Simultaneous temporal trends
in dementia incidence and prevalence, 2005‑2013:
a population-based retrospective cohort study in
Saskatchewan, Canada. Int. Psychogeriatr. 29,
1643–1658 (2016).
48. Abdulrahman, G. O. Alzheimer’s disease: current
trends in Wales. Oman Med. J. 29, 280–284
(2014).
49. Bertrand, M., Tzourio, C. & Alperovitch, A. Trends in
recognition and treatment of dementia in France
analysis of the 2004 to 2010 database of the National
Health Insurance plan. Alzheimer Dis. Assoc. Disord.
27, 213–217 (2013).
50. Ukraintseva, S., Sloan, F., Arbeev, K. & Yashin, A.
Increasing rates of dementia at time of declining
mortality from stroke. Stroke 37, 1155–1159
(2006).
51. Chien, I. C. et al. Treated prevalence and incidence of
dementia among National Health Insurance enrollees
in Taiwan, 1996–2003. J. Geriatr. Psychiatry Neurol.
21, 142–148 (2008).
52. Menec, V. H., Lix, L. & MacWilliam, L. Trends in the
health status of older Manitobans, 1985 to 1999.
Can. J. Aging 24 (Suppl. 1), 5–14 (2005).
53. Dodge, H. H. et al. Trends in the prevalence of
dementia in Japan. Int. J. Alzheimers Dis. 2012,
956354 (2012).
54. Kim, K. W. et al. A nationwide survey on the
prevalence of dementia and mild cognitive impairment
in South Korea. J. Alzheimers Dis. 23, 281–291
(2011).
55. Yu, R. et al. Trends in prevalence and mortality of
dementia in elderly Hong Kong population:
projections, disease burden, and implications for
long-term care. Int. J. Alzheimers Dis. 2012, 406852
(2012).
56. Fuh, J. & Wang, S. Dementia in Taiwan: past, present,
and future. Acta Neurol. Taiwan 17, 153–161
(2008).
57. Zhang, Y. et al. Prevalence of dementia and major
dementia subtypes in the Chinese populations: a meta-
analysis of dementia prevalence surveys, 1980–2010.
J. Clin. Neurosci. 19, 1333–1337 (2012).
58. Chan, K. Y. et al. Epidemiology of Alzheimer’s disease
and other forms of dementia in China, 1990–2010:
a systematic review and analysis. Lancet 381,
2016–2023 (2013).
59. Wu, Y.‑T., Brayne, C. & Matthews, F. E. Prevalence of
dementia in East Asia: a synthetic review of time
trends. Int. J. Geriatr. Psychiatry 30, 793–801
(2015).
60. Wu, Y.‑T. et al. Period, birth cohort and prevalence of
dementia in mainland China, Hong Kong and Taiwan:
a meta-analysis. Int. J. Geriatr. Psychiatry 29,
1212–1220 (2014).
61. Gao, M. et al. The time trends of cognitive impairment
incidence among older Chinese people in the
community: based on the CLHLS cohorts from 1998 to
2014. Age Ageing https://www.dx.doi.org/10.1093/
ageing/afx038 (2017).
62. Zeng, Y., Feng, Q., Hesketh, T., Christensen, K. &
Vaupel, J. W. Survival, disabilities in activities of daily
living, and physical and cognitive functioning among the
oldest-old in China: a cohort study. Lancet 389,
1619–1629 (2017).
63. Larson, E. B., Yaffe, K. & Langa, K. M. New insights
into the dementia epidemic. N. Engl. J. Med. 369,
2275–2277 (2013).
64. de Bruijn, R. F. et al. The potential for prevention of
dementia across two decades: the prospective,
population-based Rotterdam Study. BMC Med. 13,
132 (2015).
65. Hachinski, V. Stroke and potentially preventable
dementias proclamation. Updated World Stroke Day
proclamation. Stroke 46, 3039–3040 (2015).
66. Skoog, I. Dementia: dementia incidence — the times,
they are a‑changing. Nat. Rev. Neurol. 12, 316–318
(2016).
67. Stern, Y. Cognitive reserve. Neuropsychologia 47,
2015–2028 (2009).
68. Blazer, D. G. et al. Cognitive Aging: Progress in
Understanding and Opportunities for Action (National
Academies Press, 2015).
69. Honda, H. et al. Trends in autopsy-verified dementia
prevalence over 29 years of the Hisayama study.
Neuropathology 36, 383–387 (2016).
70. Wakutani, Y. et al. Longitudinal changes in the
prevalence of dementia in a Japanese rural area.
Psychogeriatrics 7, 150–154 (2007).
71. Jones, D. S. & Greene, J. A. Is dementia in decline?
Historical trends and future trajectories. N. Engl.
J. Med. 374, 507–509 (2016).
72. Kiyohara, Y. Epidemiology of dementia: the Hisayama
study [Japanese]. Nihon Rinsho 72, 601–606 (2014).
73. Kishimoto, H. et al. The long-term association between
physical activity and risk of dementia in the community:
the Hisayama study. Eur. J. Epidemiol. 31, 267–274
(2016).
74. Yoshitake, T. et al. Incidence and risk factors of vascular
dementia and Alzheimer’s disease in a defined elderly
Japanese population: the Hisayama study. Neurology
45, 1161–1168 (1995).
75. Ohara, T. et al. Glucose tolerance status and risk of
dementia in the community: the Hisayama study.
Neurology 77, 1126–1134 (2011).
76. Mueller, S. G. et al. Ways toward an early diagnosis in
Alzheimer’s disease: the Alzheimer’s Disease
Neuroimaging Initiative (ADNI). Alzheimers Dement. 1,
55–66 (2005).
77. Savva, G. M. et al. Age, neuropathology, and dementia.
N. Engl. J. Med. 360, 2302–2309 (2009).
78. Snowdon, D. A. Healthy aging and dementia: findings
from the Nun Study. Ann. Intern. Med. 139, 450–454
(2003).
79. Flicker, L. & Holdsworth, K. Aboriginal and Torres Strait
islander people and dementia: a review of the research.
Alzheimer’s Australia https://www.fightdementia.org.au/
files/NATIONAL/documents/Alzheimers-Australia-
Numbered-Publication-41.pdf (2014).
80. Breeze, E., Hart, N. J., Aarsland, D., Moody, C. &
Brayne, C. Harnessing the power of cohort studies for
dementia research. J. Public Mental Health 14, 8–17
(2015).
81. Rorsman, B., Hagnell, O. & Lanke, J. Prevalence and
incidence of senile and multi-infarct dementia in the
Lundby Study: a comparison between the time periods
1947–1957 and 1957–1972. Neuropsychobiology
15, 122–129 (1986).
82. Mathillas, J., Lövheim, H. & Gustafson, Y. Increasing
prevalence of dementia among very old people.
Age Ageing 40, 243–249 (2011).
83. Li, S. et al. Is the dementia rate increasing in Beijing?
Prevalence and incidence of dementia 10 years later
in an urban elderly population. Acta Psychiatr. Scand.
115, 73–79 (2007).
84. Arias, E., Heron, M. & Xu, J. United States life tables,
2012. Centers for Disease Control and Prevention
https://www.cdc.gov/nchs/data/nvsr/nvsr65/
nvsr65_08.pdf (2016).
Acknowledgment
We would like to thank Ms Lesile Grasset (Bordeaux study),
Dr Sujuan Gao (Indianapolis–Idaban Dementia Project) and
Professor Anders Wimo (Nordanstig study) for providing
age-specific and sex-specific prevalence and incidence
estimates.
Author contributions
All authors researched data for the article and reviewed
and/or edited the manuscript before submission. Y.-T.W. and
C.B. wrote the article and made substantial contributions to
­discussion of the content.
Competing interests statement
The authors declare no competing interests.
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional
claims in published maps and institutional affiliations.
Review criteria
We conducted a literature search in PubMed using search
terms (“dementia” OR “Alzheimer”) AND (“prevalence” OR
“incidence”) AND (“trend*” OR “change*”) to identify relevant
publications up to February 2017. We selected potential
population-­based studies which investigated change in preva­
lence or incidence of dementia using similar methods across
time. We also cross-checked the reference lists of previous
reports on dementia trends. We excluded studies in which
medical records or healthcare administrative databases were
used, systematic reviews, meta-analyses and those focusing
only on Alzheimer disease and cognitive impairment. We also
excluded five studies owing to methodological limita-
tions20–22,24,25. In total, we included 14 full-text studies from
Sweden, the Netherlands, Spain, France, the UK, the USA,
Japan and Nigeria, and all were published in English.
FURTHER INFORMATION
UN data: http://data.un.org/
Gapminder: http://www.gapminder.org/data/
SUPPLEMENTARY INFORMATION
See online article: S1 (table) | S2 (table) | S3 (table)
ALL LINKS ARE ACTIVE IN THE ONLINE PDF

NATURE REVIEWS | NEUROLOGY VOLUME 13 | JUNE 2017 | 339

©
2
0
1
7
M
a
c
m
i
l
l
a
n
P
u
b
l
i
s
h
e
r
s
L
i
m
i
t
e
d
,
p
a
r
t
o
f
S
p
r
i
n
g
e
r
N
a
t
u
r
e
.
A
l
l
r
i
g
h
t
s
r
e
s
e
r
v
e
d
.
REVIEWS

CORRIGENDUM

The changing prevalence and incidence of dementia over time

— current evidence
Yu‑Tzu Wu, Alexa S. Beiser, Monique M. B. Breteler, Laura Fratiglioni, Catherine Helmer, Hugh C. Hendrie,
Hiroyuki Honda, M. Arfan Ikram, Kenneth M. Langa, Antonio Lobo, Fiona E. Matthews, Tomoyuki Ohara,
Karine Pérès, Chengxuan Qiu, Sudha Seshadri, Britt-Marie Sjölund, Ingmar Skoog and Carol Brayne
Nature Reviews Neurology http://dx.doi.org/10.1038/nrneurol.2017.63 (2017)
In the version of this article initially published online, the affiliations for Karine Pérès, Chengxuan Qiu and Britt-Marie
Sjölund were incorrect. These errors have been corrected in the print, HTML and PDF versions of the article.

JUNE 2017 | VOLUME 13 www.nature.com/nrneurol

©
2
0
1
7
M
a
c
m
i
l
l
a
n
P
u
b
l
i
s
h
e
r
s
L
i
m
i
t
e
d
,
p
a
r
t
o
f
S
p
r
i
n
g
e
r
N
a
t
u
r
e
.
A
l
l
r
i
g
h
t
s
r
e
s
e
r
v
e
d
.