Available via license: CC BY-NC 4.0
Content may be subject to copyright.
Early life predictors of late life cerebral small vessel
disease in four prospective cohort studies
Ellen V Backhouse1,2, Susan D Shenkin3, Andrew M McIntosh4, Mark E Bastin1,5,6 Heather C
Whalley1,4, Maria Valdez Hernandez1,5,6, Susana Muñoz Maniega1,5,6, Mat Harris4, Aleks
Stolicyn4, Archie Campbell4, Douglas Steele7, Gordon D Waiter8, Anca-Larisa Sandu8,
Jennifer MJ Waymont5,8, Alison D Murray8, Simon R Cox9, Susanne R. de Rooij10, Tessa J.
Roseboom10, Joanna M Wardlaw1,2,5,6
1 Centre for Clinical Brain Sciences, University of Edinburgh, Chancellor’s Building, 49
Little France Crescent, Edinburgh, UK, EH16 4SB
2MRC UK Dementia Research Institute at the University of Edinburgh
3Geriatric Medicine, Usher Institute, The University of Edinburgh, 51 Little France Crescent,
Edinburgh, EH16 4SB.
4Division of Psychiatry, Royal Edinburgh Hospital, University of Edinburgh, Edinburgh, UK,
EH10 5HF
5Scottish Imaging Network, A Platform for Scientific Excellence (SINAPSE)
6Brain Research Imaging Centre, Division of Neuroimaging Sciences, Centre for Clinical
Brain Sciences, University of Edinburgh, UK
7Division of Imaging Sciences and Technology, Medical School, University of Dundee, UK,
DD19SY
8Aberdeen Biomedical Imaging Centre, School of Medicine, Medical Sciences and Nutrition,
University of Aberdeen, Foresterhill, Aberdeen AB252ZD
9Lothian Birth Cohorts Group, Department of Psychology, University of Edinburgh,
Edinburgh, UK
10Department of Epidemiology and Data Science, Amsterdam University, Medical Centres,
University of Amsterdam, The Netherlands
. CC-BY-NC 4.0 International licenseIt is made available under a
is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.(which was not certified by peer review)preprint The copyright holder for thisthis version posted September 16, 2021. ; https://doi.org/10.1101/2021.09.13.21252900doi: medRxiv preprint
NOTE: This preprint reports new research that has not been certified by peer review and should not be used to guide clinical practice.
Abstract
Development of cerebral small vessel disease, a major cause of stroke and dementia, may be
influenced by early life factors. It is unclear whether these relationships are independent of
each other, of adult socioeconomic status or of vascular risk factor exposures.
We examined associations between factors from birth (ponderal index, birth weight),
childhood (IQ, education, socioeconomic status), adult small vessel disease, and brain
volumes, using data from four prospective cohort studies: STratifying Resilience And
Depression Longitudinally (STRADL) (n=1080; mean age=59 years); The Dutch Famine
Birth cohort (n=118; mean age=68 years); the Lothian Birth Cohort 1936 (LBC1936; n=617;
mean age=73 years), and the Simpson’s cohort (n=110; mean age=78 years). We analysed
each small vessel disease feature individually and summed to give a total small vessel disease
score (range 1-4) in each cohort separately, then in meta-analysis, adjusted for vascular risk
factors and adult socioeconomic status.
Higher birth weight was associated with fewer lacunes (OR per 100g, 0.93 95%CI=0.88-
0.99), fewer infarcts (OR=0.94 95%CI=0.89-0.99), and fewer perivascular spaces (OR=0.95
95%CI=0.91-0.99). Higher childhood IQ was associated with lower white matter
hyperintensity burden (OR per IQ point=0.99 95%CI 0.98-0.998), fewer infarcts (OR=0.98,
95%CI=0.97-0.998), fewer lacunes (OR=0.98, 95%CI=0.97-0.999), and lower total small
vessel disease burden (OR=0.98, 95%CI=0.96-0.999). Low education was associated with
more microbleeds (OR=1.90 95%CI=1.33-2.72) and lower total brain volume (MD=-
178.86cm3, 95%CI=-325.07- -32.66). Low childhood socioeconomic status was associated
with fewer lacunes (OR=0.62, 95%CI=0.40-0.95).
Early life factors are associated with worse small vessel disease in later life, independent of
each other, vascular risk factors and adult socioeconomic status. Risk for small vessel disease
may originate in early life and provide a mechanistic link between early life factors and risk
of stroke and dementia. Policies investing in early child development may contribute to
improve lifelong brain health to prevent dementia and stroke in older age.
Correspondence to: Prof Joanna M. Wardlaw
Centre for Clinical Brain Sciences, University of Edinburgh, The Chancellors Building, 49
Little France Crescent, Edinburgh, EH16 4SB Email: joanna.wardlaw@ed.ac.uk Tel: 0131
537 2943
. CC-BY-NC 4.0 International licenseIt is made available under a
is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.(which was not certified by peer review)preprint The copyright holder for thisthis version posted September 16, 2021. ; https://doi.org/10.1101/2021.09.13.21252900doi: medRxiv preprint
Running title: Early life risk factors for SVD
Keywords: cerebral small vessel disease, education, childhood, MRI, epidemiology
Abbreviations: ACDS= Aberdeen Child Development Survey; ACONF= Aberdeen Children
of the 1950s cohort; DOHAD= Developmental Origins of Adult Heath and Disease;
GS:SFHS= Generation Scotland: Scottish Family Health Study; ICV= intracranial volume;
LBC1936= the Lothian Birth Cohort 1936; LGA= Lesion Growth Algorithm; PVS=
perivascular spaces; SES= socioeconomic status; STRADL= STratifying Resilience and
Depression Longitudinally; SVD= cerebral small vessel disease; WMH= white matter
hyperintensities.
. CC-BY-NC 4.0 International licenseIt is made available under a
is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.(which was not certified by peer review)preprint The copyright holder for thisthis version posted September 16, 2021. ; https://doi.org/10.1101/2021.09.13.21252900doi: medRxiv preprint
Introduction
Cerebral small vessel disease (SVD) is common at older ages1 and causes 20-25% of strokes
and up to 45% of dementias, either as vascular or mixed with Alzheimer’s disease.2 It is
responsible for up to a fifth of all strokes, doubles the risk of future stroke and worsens post-
stroke recovery.3 SVD is detected on neuroimaging or post mortem4 as white matter
hyperintensities (WMH), lacunes, microbleeds, perivascular spaces (PVS), acute lacunar
infarcts and brain atrophy.4,5 Several demographic and clinical factors are associated with
increased risk of SVD, including adult socioeconomic status (SES), hypertension and
smoking.6,7 However, a large proportion of the variance in the presence and severity of SVD
is unexplained by vascular risk factors7 and factors from earlier in life may also be
important.8
The Developmental Origins of Adult Heath and Disease (DOHAD) hypothesis9 proposes that
adverse environmental exposures occurring during gestation can cause permanent changes in
fetal development resulting in increased vulnerability to chronic diseases in adulthood.
Factors affecting foetal growth such as stress and poor nutrition10,11 are often hard to measure
but anthropometric measures such as birth weight and ponderal index (birth weight/birth
length3) can be used as proxy measures.12 Additional confounding or mediating factors in
childhood may also affect later disease risk.13 A recent meta-analysis14 found that lower
levels of childhood IQ, poorer childhood SES, and less education increased the risk of SVD
in later life by approximately 17-39%. However, it is not clear if these relationships are
independent of each other, or if they persist after adjustment for vascular risk factors and
adult SES. Few studies have examined the effect of these early life factors in combination
and many rely on childhood measures assessed retrospectively in adulthood so may be
subject to recall bias.
We examined the relationships between birth and childhood factors and total and individual
components of SVD and brain volumes, after adjustment for each other and common adult
risk factors, in four well-phenotyped prospective cohort studies: STratifying Resilience and
Depression Longitudinally (STRADL).15 the Dutch Famine Birth Cohort,16 the Lothian Birth
Cohort 1936 (LBC1936),17 and the Simpson’s cohort.18 All had information on education and
SES, and three cohorts had IQ measured during childhood. All underwent brain imaging
between the ages of 59 and 85 years. We hypothesised that low birth weight, low childhood
. CC-BY-NC 4.0 International licenseIt is made available under a
is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.(which was not certified by peer review)preprint The copyright holder for thisthis version posted September 16, 2021. ; https://doi.org/10.1101/2021.09.13.21252900doi: medRxiv preprint
IQ, low education and low childhood SES would be associated with increased SVD,
independent of each other, vascular risk factors and adult SES.
Materials and methods
Participants
The recruitment procedures and inclusion criteria for STRADL,15 the Dutch Famine Birth
Cohort,16 the LBC1936,17 and the Simpson’s cohort18 have been described previously in
detail (see Supplementary Fig. 1a-d for recruitment flow charts). All subjects were
community dwelling.
STRADL
STRADL is a population based study of 1,198 adults recruited from the Generation Scotland:
Scottish Family Health Study (GS:SFHS) and two Scottish longitudinal birth cohorts, the
Aberdeen Children of the 1950s (ACONF) cohort19 and the Walker cohort.20 ACONF
consists of surviving participants of the Aberdeen Child Development Survey (ACDS), a
population based study of schoolchildren in Aberdeen, conducted in 1962-64. The Walker
cohort is a database of over 48,000 birth records of babies born in hospital in Dundee,
between 1952 and 1966. In 2015 eligible participants were sent postal questionnaires and
between 2015 and 2019 1,188 attended in person assessments. MRI and childhood data were
available for 1080 participants (ACONF 268; Walker 201; GS:SFHS 611) (40% female;
mean age= 59.3 years, SD=10.1).
The Dutch Famine Birth cohort
The Dutch Famine Birth Cohort consists of 2,414 individuals born in the Wilhelmina
Gasthuis hospital in Amsterdam between 1st November 1943 and 28th February 1947, a
proportion of whom were exposed to the Dutch famine of 1944-1945 in utero. 151 surviving
cohort members were recruited for an MRI study in 2012 of which 118 had MRI and
childhood data (56% female; mean age=67.5 years, SD=0.9).
The Lothian Birth cohort 1936 (LBC1936)
The LBC1936 consists of 1,091 community dwelling adults born in 1936 and living in the
Lothian area of Scotland. All are surviving participants of the Scottish Mental Health Survey
1947 which was a cognitive ability test administered to all age 11 school children in Scotland
. CC-BY-NC 4.0 International licenseIt is made available under a
is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.(which was not certified by peer review)preprint The copyright holder for thisthis version posted September 16, 2021. ; https://doi.org/10.1101/2021.09.13.21252900doi: medRxiv preprint
in 1947. Between 2007 and 2009 680 of the original 1091 cohort members underwent MRI,
all with childhood data (47% female; mean age 72.7 years, SD=0.7).
The Simpson’s Cohort
The Simpson’s cohort consists of 130 individuals born 1921-1926 in three Edinburgh
hospitals. In 2000 28 people were recruited as part of the Lothian Birth Cohort 1921, 19 were
traced through hospital records from 1921 and 80 people were recruited through local
advertisements. MRI and childhood data were available for 110 people (67% female, mean
age= 78.4 years, SD=1.5).
Participants in all cohorts provided written informed consent and research was approved by
Local or Multicentre Research Ethics Committees. (STRADL:14/SS/0039; LBC1936:
MREC/01/0/56 and LREC/2003/2/29; Simpson’s cohort LREC 1702/1998/4/183–
Amendment).
Early life factors
The early life data available varied between cohorts (Fig. 1). Where possible, data were
harmonised to allow direct comparison between the studies. We examined birth weight in
grams (all cohorts) and ponderal index (birth weight/birth length3) (Dutch Famine Birth
Cohort, LBC1936 and Simpson’s cohort). In childhood, we examined: childhood IQ
(STRADL, LBC1936 and Simpson’s cohort) measured using raw test scores adjusted for age
at testing and placed on an IQ scale; education (all cohorts) dichotomised at compulsory
education (STRADL), lower secondary (Dutch Famine cohort) and 11 years (LBC1936 and
Simpson’s cohort); and childhood SES (all cohorts) classified according to parental
occupation (manual and non-manual). Further details are provided in Supplementary Table 1.
MRI acquisition and analysis
Brain imaging acquisition for STRADL,21 the Dutch Famine birth cohort,22 LBC193623 and
the Simpson’s cohort24 have been described previously. Participants were scanned on a
Philips Achieva 3.0T TX (STRADL, Aberdeen), Siemens 3T Prisma-FIT (STRADL
Dundee), a 3T Philips Ingenia (Best, the Netherlands) with a 16-channel DStream Head-Spin
coil (Dutch Famine cohort), or the same 1.5T GE Signa scanner operating in research mode
. CC-BY-NC 4.0 International licenseIt is made available under a
is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.(which was not certified by peer review)preprint The copyright holder for thisthis version posted September 16, 2021. ; https://doi.org/10.1101/2021.09.13.21252900doi: medRxiv preprint
in its original LX format (Simpson’s cohort) or following an upgrade to HDx format
(LBC1936) (Supplementary Table 2).
SVD visual ratings
Trained researchers using the same rating methods and, blind to all other data, performed all
image analyses. An experienced certified and registered neuroradiologist (JMW)
crosschecked 20% of scans. Presence of WMH, lacunes, micro-bleeds and perivascular
spaces were rated according to STRIVE criteria and established protocols, published
previously using validated visual scales,23,25,26 converted to dichotomous point scores and
summed to create the total SVD score (0-4; higher score represents higher SVD burden).6,27-29
We noted any imaging evidence of infarcts in the cortical or subcortical regions using a
validated stroke lesion rating scale.30 Superficial and deep atrophy scores were coded
separately using a valid template,31 summed to give a total score and dichotomised into ‘none
or mild’ and ‘moderate or severe’.
WMH volumes and whole brain volume
We conducted structural image analysis, blind to all non-imaging data, including
measurements of volumes of the intracranial compartment (ICV), whole brain and total
WMH volume in STRADL, LBC1936 and the Simpson’s cohort and WMH volume only in
the Dutch Famine Birth cohort. For tissue segmentation we used the processing protocol with
the Lesion Growth Algorithm (LGA), provided by the Lesion Segmentation Toolbox for
SPM (STRADL) and a semiautomatic segmentation tool MCMxxxVI previously validated32
(LBC1936 and Simpson’s cohort). We visually inspected all segmented images and manually
edited any incorrectly classified tissues. Analyses were performed using Freesurfer 5.3 and
AnalyzeTM software.
Statistical analysis
We assessed descriptive characteristics using means, standard deviations (SD), medians and
interquartile range (IQR), counts and percentages as appropriate. We used χ2 for categorical
data and Mann-Whitney U-Test for continuous data to compare differences between
participants who underwent MRI and those who did not and to examine gender differences in
SVD burden.
. CC-BY-NC 4.0 International licenseIt is made available under a
is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.(which was not certified by peer review)preprint The copyright holder for thisthis version posted September 16, 2021. ; https://doi.org/10.1101/2021.09.13.21252900doi: medRxiv preprint
Few in the cohorts had the highest SVD scores, which likely reflects the generally good
health of the cohorts. We therefore dichotomised the SVD score into 0-1 (“no or mild
disease”) and 2-4 (“moderate-severe disease”).
We performed logistic regression for differences in early life factors for higher vs lower SVD
scores and for presence of each individual SVD component and linear regression analysis to
assess early life factors and brain volumes. Brain volumes were adjusted for intracranial
volume (ICV). For all main analyses we analysed the cohorts individually and meta-analysed
them using a random effects model in Review Manager 5.3. Due to the small sample size for
some analyses we did not adjust for all available vascular risk factors. Based on previous
research6,7 we included age, sex, hypertension, smoking behaviour and adult SES at the time
of the MRI (manual vs non-manual occupation) as covariates in all models. We adjusted
analyses including birth weight and ponderal index for gestational age taken from birth
records. We performed further multiple regression analyses adjusting for the other early life
factors and where sample size allowed, using an event per variable of 10, vascular risk factors
and SES in adulthood. A Bonferroni correction for multiple testing was not appropriate, as
the variables are not independent. Therefore to mitigate the problem of multiple testing, we
defined our hypotheses a priori based on our previous meta-analysis.33
All analyses were performed using SPSS version 24 (IBM Corp., Armonk, NY) using
pairwise deletion to deal with missing data.
Data availability
The data that support the findings of this study are available upon reasonable request.
Results
Demographic and key characteristics of all participants are displayed in Table 1A-C.
Differences in demographic and key characteristics between those who underwent MRI and
those who did not are provided in the Supplementary materials and Supplementary Tables 3-
6, along with comparisons between the participants in this study and previous waves of each.
Where data were available in comparable format, we have also provided key characteristics
of the wider Scottish and Dutch population in Supplementary Tables 3-6.
Gender differences were observed in some markers of SVD. Moderate to severe SVD and
WMH burden were more common in females compared to males in the LBC1936 (SVD:
. CC-BY-NC 4.0 International licenseIt is made available under a
is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.(which was not certified by peer review)preprint The copyright holder for thisthis version posted September 16, 2021. ; https://doi.org/10.1101/2021.09.13.21252900doi: medRxiv preprint
22.4% vs 15.9%, χ2(1)= 4.7, p=0.03; WMH: 26.5% vs 18.8%; (χ2(1)= 5.8, p=0.02) and
Dutch famine birth cohort (SVD: 31.3% vs 14.0%, χ2 (1)= 4.6, p=0.03). Atrophy was more
common in males compared to females in STRADL (11.2% vs 3.1%, χ2(1)=28.0, p<0.001),
the LBC1936 (60.7% vs 40.7%, χ2(1)= 27.1, p<0.001) and Simpson’s cohort (36.4% vs
18.2%, χ2(1)= 4.2, p=0.04). No other gender differences were observed in SVD burden.
Results from our main analyses are given below. Analysis of ponderal index are detailed in
the Supplementary materials and Supplementary Fig. 2..
Birth weight
Across all four cohorts, each increase in birth weight of 100g was associated with fewer
lacunes (OR=0.93 95%CI=0.88-0.99), fewer infarcts (OR=0.94 95%CI=0.89-0.99) and
decreased moderate-severe PVS (OR=0.95 95%CI=0.91-0.99, Fig. 2A) independent of age,
sex, hypertension, smoking behaviour and adult SES. Results for the remaining lesions were
in the expected direction (increasing birth weight and lower risk of SVD features) but did not
reach significance.
Associations were attenuated but remained significant after additional adjustment for
education and childhood SES (lacunes OR=0.94, 95%CI=0.89-0.99; infarcts OR=0.94
95%CI=0.89-1.00; PVS OR=0.95 95%CI=0.91-0.99, Supplementary Table 7).
Increasing birth weight was not associated with WMH volume or brain volume in the Dutch
Famine Birth cohort, LBC1936 or Simpson’s cohort (Supplementary Table 8).
Childhood IQ
Across STRADL, LBC1936, Simpson’s, each point increase in IQ assessed in childhood was
associated with decreased risk of moderate or severe WMH (OR per point increase 0.99
95%CI=0.98-1.00), lacunes (OR=0.98 95%CI=0.97-0.99), infarcts (OR=0.98 95%CI=0.97-
1.00), and total SVD burden (OR=0.98 95%CI=0.96-1.00, Fig. 2B) independent of age, sex,
hypertension, smoking behaviour and adult SES.
Additional adjustment for education and childhood SES attenuated all associations between
childhood IQ and individual SVD features (Supplementary Table 9), but the associations with
total SVD burden (OR=0.98, 95%CI=0.97-0.997) and infarcts (OR=0.98, 95%CI=0.97-1.00,
Supplementary Table 9) remained.
. CC-BY-NC 4.0 International licenseIt is made available under a
is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.(which was not certified by peer review)preprint The copyright holder for thisthis version posted September 16, 2021. ; https://doi.org/10.1101/2021.09.13.21252900doi: medRxiv preprint
Education
Across all cohorts, low education was associated with increased risk of micro-bleeds (vs high
education level, OR=1.90 95%CI=1.33-2.72, Fig. 2C) independent of age, sex, hypertension,
smoking behaviour and adult SES. This was attenuated by additional adjustment for
childhood IQ and SES (OR=1.24, 95%CI=0.71-2.18, Supplementary Table 9) but remained.
The Simpson’s cohort were not included in this multiple regression analysis due to the small
number of participants with childhood IQ scores.
Low education was associated with lower brain volume (MD= -178.86cm3, 95%CI=-325.07-
-32.66, Supplementary Fig. 5a) but this was attenuated after adjustment for vascular risk
factors and adult SES ( = 0.01, 95%CI= -0.04-0.06, Supplementary Table 10).
Childhood SES
Across all cohorts manual childhood SES (i.e. more deprived) was associated with decreased
risk of lacunes (OR=0.62 95%CI=0.40-0.95, Fig. 2D).
Discussion
Early life factors are thought to influence health later in life but there are few studies with
such a wealth of data from birth, childhood and later life to tease out which early life factors
are important and if they are independent of each other and of exposures in later life. By
combining data from almost 2000 participants from four prospective birth cohorts we confirm
that low birth weight, low childhood IQ and less education increase SVD burden 5-8 decades
later. SVD is important since it increases dementia and stroke risk, two of the largest sources
of loss of independence, health and societal costs in older age across the world. Dementia and
stroke prevention are government priorities. Life-course models are increasingly recognised
in dementia prevention36 but have largely been ignored in stroke and SVD, which too often
focus on mid to later life only, thereby missing major opportunities to prevent these
devastating diseases much earlier, as well as gaining other health benefits.
Our findings confirm previous findings that some early life factors may increase risk of SVD
burden in later life, but importantly also demonstrate that the associations are independent of
vascular risk factors and adult SES and persist after adjustment for the other early life factors.
Lower birth weight increased the risk of lacunes, infarcts and PVS across four cohorts,
independent of education and childhood SES. In STRADL, the LBC1936 and Simpson’s
. CC-BY-NC 4.0 International licenseIt is made available under a
is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.(which was not certified by peer review)preprint The copyright holder for thisthis version posted September 16, 2021. ; https://doi.org/10.1101/2021.09.13.21252900doi: medRxiv preprint
cohort, higher childhood IQ was associated with fewer infarcts and lacunes, lower WMH and
total SVD burden. Associations between childhood IQ, infarcts and total SVD burden were
independent of education and childhood SES. Across all cohorts, low education level was
associated with more micro-bleeds. These new data show that lower birth weight, childhood
IQ and low education are independently associated with increased SVD lesions many decades
later.
Low childhood SES was not found to be associated with SVD and associations between
childhood SES and lacunes were in the opposite direction to what we expected. This was true
for univariate analyses (Supplementary Table 8) and multivariate analyses. This may be
because childhood SES reflects SES in adulthood, whereas the other early life factors such as
cognitive ability and education capture different aspects of early life adversity. Alternatively,
parental occupation, which we used as a measure of SES to allow direct comparison between
cohorts, may not have been a sufficiently sensitive measure of actual SES in childhood. Jobs
traditionally classed as ‘manual’ such as farmer or skipper trawler can have a high income
and the wartime occupations of the parents of some cohort members would have been
limited. In the LBC1936 we have previously shown a trend towards an association between
SVD at age 72 and age 11 deprivation index.37,38 Deprivation index encompasses several
socioeconomic markers so may be a better measure of SES and thus of associations with
SVD in later life.
Increasing age and traditional vascular risk factors, particularly hypertension, are important
risk factors for SVD1,39 but together explain little variance in WMH (~2%)40,41 suggesting
that other factors, as identified here, may contribute to SVD pathology. The effect sizes are
small when considered per point difference in IQ score or per 100g difference in birth weight,
and the early life variables examined here only explained ~1% of the variance in SVD risk.
However, the fact that these effects are evident for such small differences in scores or
weights, and at up to seven decades later, underscores that factors influencing early stages in
life, including during foetal development and childhood, can impact on brain health in older
age and are rightly public health priorities. Furthermore it is likely that our effects are an
underestimate of population effects given that our cohorts are healthier with higher IQ than
average members of the population. For example the mean age 11 IQ score of the LBC1936
was relatively high with a narrow range compared with the mean age 11 IQ for Scotland in
1947.42
. CC-BY-NC 4.0 International licenseIt is made available under a
is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.(which was not certified by peer review)preprint The copyright holder for thisthis version posted September 16, 2021. ; https://doi.org/10.1101/2021.09.13.21252900doi: medRxiv preprint
Our associations between birth weight and SVD are independent of gestational age and
therefore reflect the impact of variations in growth rather than prematurity. The relationship
between size at birth and brain structure is biologically plausible: lack of nutrients at
particular stages of gestation can impair foetal growth resulting in small size at birth,
indicated by low birth weight or disproportionate growth such as low birth weight to length
ratio (ponderal index). Long lasting physiological changes in the structure of foetal organs
and tissues can increase risk of later disease in adulthood.43,44 Relations between size at birth
and disease in later life including coronary heart disease (CHD)45,46 are well established, but
fewer studies have examined brain health, particularly with this sample size or age range. The
current study is one of the few examining the effect of size at birth on brain volumes in later
life and the first to examine multiple markers of SVD.
We found no associations between birth weight or ponderal index and WMH burden or brain
volumes. This is consistent with data from the (AGES)-Reykjavik study (RS)47 which
reported no association between ponderal index and WMH burden at age 75 after adjustment
for vascular risk factors. Birth weight and size are indirect measures of the fetal environment
and may not reflect all adverse prenatal circumstances that can affect later life health. The
Dutch Famine Birth Cohort previously showed that foetal malnutrition can lead to accelerated
cognitive ageing and advanced structural brain ageing, measured using the BrainAGE method
(a composite measure based mainly on tissue loss) independent of birth weight.48
From a life course perspective, a disadvantaged foetal environment may interact with factors
during childhood to increase risk of later disease. Development of neural pathways in the
brain extends well into childhood and may therefore mean the brain remains vulnerable to
insults for a longer period of time.49 Our two recent meta-analyses14,50 found small but
statistically significant associations between increasing childhood IQ and lower WMH
burden (r = -0.07) and a 17% lower risk of stroke. Low education (defined by attainment or
years) was associated with a 35% relative increased risk of stroke and a 17% increased risk of
SVD. Manual paternal occupation (SES measure) was associated with a 28% increased risk
of stroke and increased WMH (only one study identified). However, the previous literature
did not allow us to determine the independent effect of these three inter-related early life
factors from each other, or from risk factor exposures in adulthood, which we are now able to
do.
. CC-BY-NC 4.0 International licenseIt is made available under a
is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.(which was not certified by peer review)preprint The copyright holder for thisthis version posted September 16, 2021. ; https://doi.org/10.1101/2021.09.13.21252900doi: medRxiv preprint
In many high-income countries age specific incidence rates of dementia are declining.51,36
Improved health in old age, including cerebrovascular disease and SVD52, has been reported
across generations and epidemiological studies have found that age adjusted incidence rates
of dementia are lower in more recent cohorts compared with cohorts from previous
decades.36,51,52 This can in part be attributed to population public health strategies, advances
in treatment and management of patients with cerebrovascular disease and dementia, and
improved management of key modifiable risk factors such as smoking and hypertension.
Additionally, investment in early life, particularly improvements in living conditions and
education, explain some of the decline in incidence of dementia.53-55 More recent generations
of older adults have received more years of statutory education than older cohorts which may
increase cognitive reserve and therefore reduce risk of dementia or cerebrovascular disease.
This is particularly relevant to our cohorts, whose years of birth span the 20th century. Low
education increased with increasing age of our cohorts, as did SVD burden. In STRADL
(median year of birth 1955) 24% had low education and 17.3% had moderate to severe SVD
burden. In the Simpson’s cohort (born 1921-1926) 81% had low education and 32% had
moderate to severe SVD burden. Increases in life expectancy means that the global
population is aging, therefore identifying factors that contribute to reductions in the
prevalence and incidence of dementia and cerebrovascular disease is a major priority. Our
findings support the suggestion that reducing inequalities, including improvements in
education, will contribute to improvements in health in older age and a reduction in the risk
of dementia and cerebrovascular disease.
Why might the early life factors increase the risk of SVD in later life? There are numerous
potential explanations. Children with higher IQ or from higher socioeconomic backgrounds
are likely to receive better diets, medical care, more educational opportunities and hence
better job opportunities or less hazardous working conditions. In adulthood, they may be
more likely to engage in better lifestyle behaviours and self-management of vascular risk
factors. Alternatively, positive early life factors may be associated with, or lead to, an
increase in the resilience and integrity of the brain resulting in less SVD. These remain
important empirical questions to be addressed in future work.
Strengths and limitations
Strengths include data collected prospectively in early life through to middle or later life,
including brain imaging, from different studies in two western European countries. Detailed
. CC-BY-NC 4.0 International licenseIt is made available under a
is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.(which was not certified by peer review)preprint The copyright holder for thisthis version posted September 16, 2021. ; https://doi.org/10.1101/2021.09.13.21252900doi: medRxiv preprint
birth records allowed correction for gestational age and did not rely on retrospective
estimations of birth weight. We used ponderal index and birth weight as measures of infant
growth. Ponderal index may be a better indicator of gestational problems than birth-weight
percentiles as it provides information on the neonate’s body proportionality and can detect
situations in which weight growth exceeds or fails to match growth in the infant’s length.56
We adjusted for key adult vascular risk factors and other early life factors in our analyses
with a relatively large sample size for some analyses. We also did a detailed characterisation
of SVD using multiple individual assessments as well as a summary score.
Limitations include availability of birth data only for some participants in STRADL and the
LBC1936. Participants in the Dutch Famine Birth Cohort may be unusual due to their famine
exposure, and we have demonstrated excess mortality up to the age of 63 years in women
exposed to famine in early gestation57. This may have resulted in selective participation of
people who were in sufficient health to participate in the present study at age 68 years.
Participants with birth data were born in hospitals, which was uncommon at the time of their
births. In the Netherlands women largely delivered at home supported by a midwife. Whilst
little is known about the actual referring pattern during this period most referrals to hospital
were because of social or medical reasons and most referred women were from lower or
middle social classes. Two of our cohort’s early childhood or early adulthood were spent during the
Second World War, which may have influenced the development of cognitive ability or educational
opportunities. Although this seems unlikely as IQ scores of those who took the Moray House Test No.
12 in 1947 (born 1936) were higher than the cohort who took them in 1932 (born 1921). The four
cohorts recruited community-dwelling volunteers who may be healthier, with less
socioeconomic adversity than non-volunteers. Within our cohorts those who completed the
MRI were younger and healthier than those who declined. Participants in all but one cohort
were largely female and when compared to the Scottish and Dutch population had lower risk
factor profiles, were more educated and from higher adult socioeconomic class. Even in our
oldest cohort aged 80 years, less than 30% of participants had moderate or severe SVD. The
large sample size of some of our cohorts mean that there are participants with a range of
socioeconomic backgrounds and medical conditions, but our samples may not be truly
representative of the populations from which they are drawn. Our samples came from three
regions of Scotland and one region of the Netherlands, which may introduce effects due to
local variations in socioeconomic strata but may also increase the generalisability of our
findings and may also be considered a strength of our study. Years of education were not
. CC-BY-NC 4.0 International licenseIt is made available under a
is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.(which was not certified by peer review)preprint The copyright holder for thisthis version posted September 16, 2021. ; https://doi.org/10.1101/2021.09.13.21252900doi: medRxiv preprint
available for all cohorts and the education system in the Netherlands differs from that in
Scotland which meant the division into ‘low’ and ‘high’ education level was relatively crude.
Whilst we adjusted our models for key vascular risk factors, it was not possible to separate
the confounding effects of other prenatal environmental or genetic influences which may
affect foetal brain development. In this study we did not adjust for multiple comparisons as a
Bonferroni-style correction would have been inappropriate when our variables are not
independent. We dealt with multiple comparisons as recommended by Perneger58 by
transparently reporting all results, including those with borderline significance. We also
specified our hypotheses a priori based on previous research. However, given the number of
statistical comparisons in our analysis it is still possible that some of our associations may be
due to Type I error.
SVD frequently coesixtis with neurodegenerative disease. We did not examine associations
between early life factors and biomarkers such as amyloid-β, tau or synclein but given the
overlap between neurodegenerative and cerebrovascular pathologies, including shared risk
factors59, it is possible that the associations observed here may interact with degenerative
neuropathologies.
Conclusions
Our findings suggest an important effect of early life factors, particularly childhood IQ, on
brain vascular disease in later life, independent of common vascular risk factors, adult SES
and other early life factors. Positive early life factors may influence health behaviours and
access to socioeconomic resources beneficial to health, or may increase brain integrity and
resilience reducing susceptibility to cerebrovascular disease. Brain vascular disease increases
the risk of cognitive impairment, dementia and stroke1 and worsens chances of recovery after
stroke.3 The current findings may provide a possible mechanistic link between early life
factors and risk of stroke and dementia. Health disparities are well known and these findings
suggest that such disparities may have effects persist across more than seven decades of life,
highlighting the importance of identifying modifiable early life factors as targets for future
social policy interventions with have long-lasting impacts.
Funding
Generation Scotland received core support from the Chief Scientist Office of the Scottish
Government Health Directorates [CZD/16/6] and the Scottish Funding Council [HR03006]
. CC-BY-NC 4.0 International licenseIt is made available under a
is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.(which was not certified by peer review)preprint The copyright holder for thisthis version posted September 16, 2021. ; https://doi.org/10.1101/2021.09.13.21252900doi: medRxiv preprint
and is currently supported by the Wellcome Trust [216767/Z/19/Z]. The MRI data collection
was funded by the Wellcome Trust (Wellcome Trust Strategic Award “STratifying Resilience
and Depression Longitudinally” (STRADL) Reference 104036/Z/14/Z).” The LBC1936 is
supported by Age UK [MR/M01311/1] (http://www.disconnectedmind.ed.ac.uk) and the
Medical Research Council [G1001245/96099]. LBC1936 MRI brain imaging was supported
by Medical Research Council (MRC) grants [G0701120], [G1001245], [MR/M013111/1] and
[MR/R024065/1] and Row Fogo Charitable Trust (Grant No. BROD.FID3668413).
Simpson’s Cohort was supported by the UK MRC and Chest Heart Stroke Scotland. JMJW
received funding from TauRx Pharmaceuticals Ltd. EB received funding from the Sackler
Foundation. JMW received funding from the UK Dementia Research Institute (DRI Ltd,
funded by the UK Medical Research Council, Alzheimer’s Society and Alzheimer’s Research
UK) and SVDs@Target, the Fondation Leducq Transatlantic Network of Excellence for the
Study of Perivascular Spaces in Small Vessel Disease, ref no. 16 CVD 05.
Competing interests
The authors report no competing interests.
References
1. Debette S, Schilling S, Duperron M-G, Larsson S, Markus H. Clinical significance of
magnetic resonance imaging markers of vascular brain injury: A systematic review
and meta-analysis. JAMA Neurol. 2019;76:81-94
2. Hachinski V, Einhäupl K, Ganten D, al. e. Preventing dementia by preventing stroke:
The berlin manifesto. Alzheimers Dement. 2019;15:961-984
3. Georgakis MK, Duering M, Wardlaw JM, Dichgans M. WMH and long-term
outcomes in ischemic stroke. A systematic review and meta-analysis. Neurology.
2019;92:e1298-e1308
4. Wardlaw JM, Smith C, Dichgans M. Mechanisms of sporadic cerebral small vessel
disease: Insights from neuroimaging. Lancet Neurol. 2013;12:483-497
5. Wardlaw JM, Smith EE, Biessels GJ et al. Neuroimaging standards for research into
small vessel disease and its contribution to ageing and neurodegeneration. Lancet
Neurol. 2013;12:822-838
. CC-BY-NC 4.0 International licenseIt is made available under a
is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.(which was not certified by peer review)preprint The copyright holder for thisthis version posted September 16, 2021. ; https://doi.org/10.1101/2021.09.13.21252900doi: medRxiv preprint
6. Staals J, Makin SDJ, Doubal FN, Dennis MS, Wardlaw JM. Stroke subtype, vascular
risk factors, and total mri brain small-vessel disease burden. Neurology.
2014;83:1228-1234
7. Wardlaw JM, Allerhand M, Doubal FN et al. Vascular risk factors, large-artery
atheroma, and brain white matter hyperintensities. Neurology. 2014;82:1331-1338
8. Backhouse EV, McHutchison CA, Cvoro V, Shenkin SD, Wardlaw JM. Early life risk
factors for stroke and cognitive impairment. Curr Epidemiol Rep. 2015;2:172-179
9. Barker D. Developmental origins of adult health and disease. J Epidemiol Comm
Health. 2004;58:114-115
10. Marques A, O'Connor T, Roth C, Susser E, Bjorke-Monsen A. The influence of
maternal prenatal and early childhood nutrition and maternal prenatal stress on
offspring immune system development and neurodevelopmental disorders. Front
Neurosci. 2013;7:120
11. Franke K, Van den Bergh BRH, de Rooij SR, Kroegel N, Nathanielsz PW, Rakers F,
et al. Effects of maternal stress and nutrient restriction during gestation on offspring
neuroanatomy in humans. Neurosci Biobehav Rev. 2020
12. Wilcox A. On the importance -and the unimportanceof birthweight. Int J Epidemiol.
2001;30:1233-1241
13. Hamilton OK, Backhouse EV, Janssen E et al. Cognitive impairment in sporadic
cerebral small vessel disease: A systematic review and meta-analysis. Alzheimers
Dement.2020; Nov13. doi: 10.1002/alz.12221
14. Backhouse E, McHutchison C, Cvoro V, Shenkin S, Wardlaw J. Early life risk factors
for cerebrovascular disease: A systematic review and meta-analysis. Neurology.
2017;88:976-984
15. Navrady LB, Wolters MK, MacIntyre DJ et al. Cohort profile: Stratifying resilience
and depression longitudinally (STRADL): a questionnaire follow-up of generation
scotland: Scottish family health study (GS:SFHS). Int J Epidemiol. 2018;47:13-14g
16. Ravelli ACJ, van der Meulen JHP, Michels RPJ et al. Glucose tolerance in adults after
prenatal exposure to famine. Lancet. 1998;351:173-177
17. Taylor AM, Pattie A, Deary IJ. Cohort profile update: The Lothian Birth Cohorts of
1921 and 1936. Int J Epidemiol. 2018;47(4):1042-1042r
18. Shenkin S, Bastin M, MacGillivray T, Deary I, Starr J, Wardlaw J. Birth parameters
are associated with late-life white matter integrity in community-dwelling older
people. Stroke. 2009;40:1225-1228
. CC-BY-NC 4.0 International licenseIt is made available under a
is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.(which was not certified by peer review)preprint The copyright holder for thisthis version posted September 16, 2021. ; https://doi.org/10.1101/2021.09.13.21252900doi: medRxiv preprint
19. Leon DA, Lawlor DA, Clark H, Macintyre S. Cohort profile: The Aberdeen Children
of the 1950s study. Int J Epidemiol. 2006;35:549-552
20. Libby G, Smith A, McEwan NF et al. The walker project: a longitudinal study of
48,000 children born 1952-1966 (aged 36-50 years in 2002) and their families.
Paediatr Perinat Epidemiol. 2004;18:302-312
21. Habota T, Sandu A, Waiter G et al. Cohort profile for the stratifying resilience and
depression longitudinally (STRADL) study: a depression-focused investigation of
generation scotland, using detailed clinical, cognitive, and neuroimaging assessments.
Wellcome Open Research. 2019;4,185
22. de Rooij SR, Caan MWA, Swaab DF et al. Prenatal famine exposure has sex-specific
effects on brain size. Brain. 2016;139:2136-2142
23. Doubal FN, Dennis M, Wardlaw J. Characteristics of patients with minor ischaemic
strokes and negative MRI: A cross-sectional study. J Neurol Neurosurg Psychiatry.
2011;82:540-542
24. Shenkin SD, Bastin ME, MacGillivray TJ et al. Cognitive correlates of cerebral white
matter lesions and water diffusion tensor parameters in community-dwelling older
people. Cerebrovasc Dis. 2005;20:310-318
25. Cordonnier C, Potter GM, Jackson CA et al. Improving interrater agreement about
brain microbleeds. Development of the Brain Observer MicroBleed Scale (BOMBS).
Stroke 2009;49:94-99
26. Potter GM, Chappell FM, Morris Z, Wardlaw JM. Cerebral perivascular spaces
visible on magnetic resonance imaging: Development of a qualitative rating scale and
its observer reliability. Cerebrovasc Dis. 2015;39:224-231
27. Klarenbeek P, van Oostenbrugge RJ, Rouhl RPW, Knottnerus ILH, Staals J.
Ambulatory blood pressure in patients with lacunar stroke. Association With Total
MRI Burden of Cerebral Small Vessel Disease. Stroke. 2013;44:2995-2999
28. Huijts M, Duits A, Van Oostenbrugge R, Kroon A, De Leeuw P, Staals J.
Accumulation of MRI markers of cerebral small vessel disease is associated with
decreased cognitive function. A study in first-ever lacunar stroke and hypertensive
patients. Front Aging Neurosci. 2013;5:72
29. Staals J, Booth T, Morris Z et al. Total mri load of cerebral small vessel disease and
cognitive ability in older people. Neurobiol Aging. 2015;36:2806-2811
30. Wardlaw JM, Sellar R. A simple practical classification of cerebral infarcts on CT and
its interobserver reliability. Am J Neuroradiol. 1994;15:1933-1939.
. CC-BY-NC 4.0 International licenseIt is made available under a
is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.(which was not certified by peer review)preprint The copyright holder for thisthis version posted September 16, 2021. ; https://doi.org/10.1101/2021.09.13.21252900doi: medRxiv preprint
31. Farrell C, Chappell F, Armitage PA et al. Development and initial testing of normal
reference MR images for the brain at ages 65–70 and 75–80 years. Eur Radiol.
2009;19:177-183
32. Aribisala B, Valdés Hernández M, Royle N et al. Brain atrophy associations with
white matter lesions in the ageing brain: The Lothian Birth Cohort 1936. Eur Radiol.
2013;23:1084-1092
33. Bender R., Bunce C., Clarke M et al. Attention should be given to multiplicity issues
in systematic reviews. J Clin Epidemiol. 2008;61:857-865
34. Fazekas F, Chawluk J, Alavi A, Hurtig H, Zimmerman R. Mr signal abnormalities at
1.5T in Alzheimer's dementia and normal aging. Am J Roentgenol. 1987;149:351-356
35. Doubal F, MacLullich A, Ferguson K, Dennis M, Wardlaw J. Enlarged perivascular
spaces on mri are a feature of cerebral small vessel disease. Stroke. 2010;41:450-454
36. Livingston G, Huntley J, Sommerlad A et al. Dementia prevention, intervention, and
care: 2020 report of the Lancet commission. Lancet. 2020;396:413-446
37. Field T, Doubal F, Johnson W et al. Early life characteristics and late life burden of
cerebral small vessel disease in the lothian birth cohort 1936. Aging. 2016;8:2039
38. Valdés Hernández M, Booth T, Murray C et al. Brain white matter damage in aging
and cognitive ability in youth and older age. Neurobiol Aging. 2013;34:2740-2747
39. Dufouil C, de Kersaint–Gilly A, Besancon V et al. Longitudinal study of blood
pressure and white matter hyperintensities the eva mri cohort. Neurology.
2001;56:921-926
40. Wardlaw J, Allerhand M, Doubal F et al. Vascular risk factors, large-artery atheroma,
and brain white matter hyperintensities. Neurology. 2014;82:1331-1338
41. Cox S, Lyall D, Ritchie S et al. Associations between vascular risk factors and brain
MRI indices in UK biobank. Eur Heart J. 2019;40:2290-2300
42. Murray C., Johnson W., Wolf MS., IJ. D. The association between cognitive ability
across the lifespan and health literacy in old age: The Lothian Birth Cohort 1936.
Intelligence. 2011;39:178-187
43. Langley Evans S. Nutritional programming of disease: Unravelling the mechanism.
J Anatomy. 2009;215:36-51
44. Goyal D., Limesand SW., R. G. Epigenetic responses and the developmental origins
of health and disease. J Endocrinol. 2019;242:T105
. CC-BY-NC 4.0 International licenseIt is made available under a
is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.(which was not certified by peer review)preprint The copyright holder for thisthis version posted September 16, 2021. ; https://doi.org/10.1101/2021.09.13.21252900doi: medRxiv preprint
45. Smith CJ, Ryckman KK, Barnabei VM, Howard BV, Isasi CR, Sarto GE, et al. The
impact of birth weight on cardiovascular disease risk in the women's health initiative.
Nutrition, Metabolism and Cardiovascular Diseases. 2016;26:239-245
46. Lawlor D, Ronalds G, Clark H, Davey Smith G, Leon D. Birth weight is inversely
associated with incident coronary heart disease and stroke among individuals born in
the 1950s: findings from the Aberdeen Children of the 1950s prospective cohort
study. Circulation. 2005;112:1414-1418
47. Muller M, Sigurdsson S, Kjartansson O et al. Birth size and brain function 75 years
later. Pediatrics. 2014;134:761-770
48. de Rooij S, Wouters H, Yonker J, Painter R, Roseboom T. Prenatal undernutrition and
cognitive function in late adulthood. Prod Natl Acad Sci. 2010;107:16881-16886
49. Walker S, Wachs T, Grantham-McGregor S et al. Inequality in early childhood: risk
and protective factors for early child development. Lancet. 2011;378:1325-1338
50. McHutchison C, Backhouse E, Cvoro V, Shenkin S, Wardlaw J. Education,
socioeconomic status, and intelligence in childhood and stroke risk in later life: A
meta-analysis. Epidemiology. 2017;28:608-618
51. Wu Wu YT, Beiser AS, Breteler MM, Fratiglioni L, Helmer C, Hendrie HC, Honda
H, Ikram MA, Langa KM, Lobo A, Matthews FE. The changing prevalence and
incidence of dementia over time—current evidence. Nat Rev Neurol. 2017. 13(6):327.
52. Schrijvers, E. M. C. et al. Is dementia incidence declining? Trends in dementia
incidence since 1990 in the Rotterdam Study. Neurology (2012) 78, 1456–1463.
53. 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 2013. 80, 1888–1894.
54. Satizabal, C. L. et al. Incidence of dementia over three decades in the Framingham
Heart Study. N. Engl. J. Med 2016. 374, 523–532.
55. Langa, K. et al. A comparison of the prevalence of dementia in the United States in
2000 and 2012. JAMA Intern. Med. 2017. 177, 51–58.
56. Patterson R, Pouliot M. Neonatal morphometrics and perinatal outcome: Who is
growth retarded? Am J Obstet Gynecol. 1987;157:691-693
57. van Abeelen AF, Veenendaal MV, Painter RC et al. Survival effects of prenatal
famine exposure. Am J Clin Nutr. 2012;95:179-183
. CC-BY-NC 4.0 International licenseIt is made available under a
is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.(which was not certified by peer review)preprint The copyright holder for thisthis version posted September 16, 2021. ; https://doi.org/10.1101/2021.09.13.21252900doi: medRxiv preprint
58. Perneger TV. What's wrong with Bonferroni adjustments. Bmj. (1998)
18;316(7139):1236-8.
59. Greenberg SM, Bacskai BJ, Hernandez-Guillamon M, Pruzin J, Sperling R, van
Veluw SJ. Cerebral amyloid angiopathy and Alzheimer disease—one peptide, two
pathways. Nature Reviews Neurology. 2020 Jan;16(1):30-42.
60. Gray L., Leyland AH., Currie E., Montagu I., Bardsley D., Terje A. et al (2016)
Scottish Health Survey 2016, Volume 1- main report.
https://www.gov.scot/publications/scottish-health-survey-2016-volume-1-main-
report/pages/47/
61. 61 Office for National Statistics. (2020). 2011 Census: Aggregate Data. [data
collection]. UK Data Service. SN: 7427, http://doi.org/10.5257/census/aggregate-
2011-2
62. Smith BH, Campbell A, Linksted P, et al.: Cohort profile: Generation Scotland:
Scottish Family Health Study (GS:SFHS). The study, its participants and their
potential for genetic research on health and illness. Int J Epidemiol. 2013; 42(3): 689–
700.
63. Statistics Netherlands (2014) Dutch Census 2011. Central Bureau of statistics
Netherlands. https://www.cbs.nl/en-gb/publication/2014/47/dutch-census-2011
64. Schelleman, H., Klungel, O., Kromhout, D. et al. Prevalence and determinants of
undertreatment of hypertension in the Netherlands. J Hum Hypertens 18, 317–324
(2004). https://doi.org/10.1038/sj.jhh.1001672
65. Smoking, alcohol and drugs: Health Survey/Lifestyle Monitor, Statistics Netherlands
(CBS) in collaboration with National Institute for Public Health and the Environment
(RIVM) and Trimbos-institute, 20xx
66. Zorlu A. Occupational adjustment of immigrants in the Netherlands. Journal of
International Migration and Integration. 2013 14(4):711-31.
67. Ravelli, A. C. J. (1999). Prenatal exposure to the Dutch famine and glucose tolerance
and obesity at age 50. Universiteit van Amsterdam.
68. Painter RC, de Rooij SR, Bossuyt PM, et al.. Blood pressure response to
psychological stressors in adults after prenatal exposure to the Dutch famine. Journal
of Hypertension. 2006;24(9):1771–1778. doi: 10.1097/01.hjh.0000242401.45591.e7.
69. Bromly C., Sproston k., Shelton N. (2004) Scottish Health Survey 2003 summary of
findings, The National Centre for Social Research and Department of Epidemiology
and Pubilic Health, University College London
https://www.webarchive.org.uk/wayback/archive/20141129163333/http://www.scotla
nd.gov.uk/Publications/2005/12/02160336/03367
. CC-BY-NC 4.0 International licenseIt is made available under a
is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.(which was not certified by peer review)preprint The copyright holder for thisthis version posted September 16, 2021. ; https://doi.org/10.1101/2021.09.13.21252900doi: medRxiv preprint
Figure legends
Figure 1: The lifecourse perspective of the risk of SVD and stroke. Adapted from Figure
1, Backhouse Curr Epidemiol Rep 2015 2:172-179.
Figure 2 (A-D) Forest plots showing associations between features of SVD and (a) birth
weight (b) childhood IQ (c) Low education (d) Low childhood SES. All analyses are
adjusted for age, sex, hypertension, smoking behaviour and adult SES.
. CC-BY-NC 4.0 International licenseIt is made available under a
is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.(which was not certified by peer review)preprint The copyright holder for thisthis version posted September 16, 2021. ; https://doi.org/10.1101/2021.09.13.21252900doi: medRxiv preprint
. CC-BY-NC 4.0 International licenseIt is made available under a
is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.(which was not certified by peer review)preprint The copyright holder for thisthis version posted September 16, 2021. ; https://doi.org/10.1101/2021.09.13.21252900doi: medRxiv preprint
. CC-BY-NC 4.0 International licenseIt is made available under a
is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.(which was not certified by peer review)preprint The copyright holder for thisthis version posted September 16, 2021. ; https://doi.org/10.1101/2021.09.13.21252900doi: medRxiv preprint
. CC-BY-NC 4.0 International licenseIt is made available under a
is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.(which was not certified by peer review)preprint The copyright holder for thisthis version posted September 16, 2021. ; https://doi.org/10.1101/2021.09.13.21252900doi: medRxiv preprint
. CC-BY-NC 4.0 International licenseIt is made available under a
is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.(which was not certified by peer review)preprint The copyright holder for thisthis version posted September 16, 2021. ; https://doi.org/10.1101/2021.09.13.21252900doi: medRxiv preprint
