ARTICLE OPEN ACCESS
Genetically determined blood pressure,
antihypertensive drug classes, and risk of
stroke subtypes
Marios K. Georgakis, MD, PhD,* Dipender Gill, MD, PhD,* Alastair J.S. Webb, DPhil, Evangelos Evangelou, PhD,
Paul Elliott, PhD, Cathie L.M. Sudlow, DPhil, Abbas Dehghan, MD, PhD, Rainer Malik, PhD,
Ioanna Tzoulaki, PhD,†and Martin Dichgans, MD†
Neurology®2020;95:1-e9. doi:10.1212/WNL.0000000000009814
Correspondence
Dr. Dichgans
martin.dichgans@
med.uni-muenchen.de
Abstract
Objective
We employed Mendelian randomization to explore whether the effects of blood pressure (BP)
and BP-lowering through different antihypertensive drug classes on stroke risk vary by stroke
etiology.
Methods
We selected genetic variants associated with systolic and diastolic BP and BP-lowering variants
in genes encoding antihypertensive drug targets from genome-wide association studies
(GWAS) on 757,601 individuals. Applying 2-sample Mendelian randomization, we examined
associations with any stroke (67,162 cases; 454,450 controls), ischemic stroke and its subtypes
(large artery, cardioembolic, small vessel stroke), intracerebral hemorrhage (ICH, deep and
lobar), and the related small vessel disease phenotype of white matter hyperintensities (WMH).
Results
Genetic predisposition to higher systolic and diastolic BP was associated with higher risk of any
stroke, ischemic stroke, and ICH. We found associations between genetically determined BP
and all ischemic stroke subtypes with a higher risk of large artery and small vessel stroke
compared to cardioembolic stroke, as well as associations with deep, but not lobar ICH. Genetic
proxies for calcium channel blockers, but not β-blockers, were associated with lower risk of any
stroke and ischemic stroke. Proxies for calcium channel blockers showed particularly strong
associations with small vessel stroke and the related radiologic phenotype of WMH.
Conclusions
This study supports a causal role of hypertension in all major stroke subtypes except lobar ICH.
We find differences in the effects of BP and BP-lowering through antihypertensive drug classes
between stroke subtypes and identify calcium channel blockade as a promising strategy for
preventing manifestations of cerebral small vessel disease.
*These authors contributed equally to this work as co–first authors.
†These authors contributed equally to this work as co–last authors.
From the Institute for Stroke and Dementia Res earch (ISD), University Hospital (M. K.G., R.M., M.D.), and Graduate School for Sy stemic Neurosciences (M.K.G.), Ludwig-Max imilians-
Universit¨
at LMU, Munich, Germany; Department of Biosta tistics and Epidemiology, School of Public Health (D.G., E.E., C.L.M.S., A.D., I.T. ), UK Dementia Research Institute (P.E., A.D.),
Health Data Research-UK London (P. E.), and MRC-PHE Centre for Environment, School of Publ ic Health (I.T.), Imperial College London; Centre for Pr evention of Stroke and Dementia,
Department of Clinical Neurosciences (A.J.S.W .), University of Oxford, UK; Department of Hygiene and Epidemiology (E.E., I.T.), Uni versity of Ioannina Medical School, Greece; National
Institute for Health Research Imperia l College Biomedical Research Centre (P.E.), London; Institute for Genetics and Molecular Medicine (C.L.M.S.), University of Ed inburgh, UK; Munich
Cluster for Systems Neurology (SyNergy) (M. D.); and German Centre for Neurodegenerative Diseas es (DZNE) (M.D.), Munich, Germany.
Go to Neurology.org/N for full disclosures. Funding information and disclosures deemed relevant by the authors, if any, are provided at the end of the article.
The Article Processing Charge was funded by Imperial College London.
This is an open access article distributed under the Creative Commons Attribution License 4.0 (CCBY), which permits unrestricted use, distribution, and reproduction in any medium,
provided the original work is properly cited.
Copyright © 2020 The Author(s). Published by Wolters Kluwer Health, Inc. on behalf of the American Academy of Neurology. 1
Published Ahead of Print on July 1, 2020 as 10.1212/WNL.0000000000009814
Stroke ranks among the leading causes of death and disability
worldwide.
1,2
High blood pressure (BP) is the major risk
factor for both ischemic and hemorrhagic stroke, accounting
for ;50% of the population attributable risk worldwide.
3–6
BP lowering reduces stroke risk with known differences be-
tween antihypertensive drug classes.
7,8
Randomized con-
trolled trials (RCTs) found calcium channel blockers (CCBs)
to be superior to other drug classes, and specifically β-blockers
(BB), in lowering stroke risk.
7,9,10
However, it remains un-
known whether the effects of BP or BP lowering through
specific drug classes vary between stroke etiologies. In light of
largely variable mechanisms between large artery stroke
(LAS), cardioembolic stroke (CES), small vessel stroke
(SVS), and deep and lobar intracerebral hemorrhage
(ICH),
11,12
differences seem possible and might have rele-
vance for therapeutic decisions.
Mendelian randomization uses genetic variants as proxies
for traits of interest and is by design less prone to con-
founding and reverse causation than observational stud-
ies.
13
As such, Mendelian randomization has been proven
valuable in exploring causality and in predicting the effects
of interventions,
13–17
as we recently showed for the effects
of antihypertensive drugs on vascular outcomes.
18
The
large samples in genome-wide association studies (GWAS)
further permit exploration of outcomes for which there are
no adequate data from RCTs, as is the case for BP-lowering
and stroke subtypes. Here, leveraging genetic data on BP
19
and stroke,
20
we employed Mendelian randomization to
examine the effects of genetically determined BP and ge-
netic proxies for antihypertensive drug classes on stroke
subtypes, as well as on white matter hyperintensities
(WMH), a radiologic manifestation of small vessel dis-
ease (SVD).
Methods
Standard protocol approvals, registrations,
and patient consents
This study was conducted in accordance with the guidelines
for Strengthening the Reporting of Observational Studies in
Epidemiology–Mendelian randomization (STROBE-MR).
21
All data were derived from studies that had already obtained
ethical review board approvals.
Genetic instrument selection
Data sources are detailed in table 1. We used summary sta-
tistics from the discovery GWAS meta-analysis of the In-
ternational Consortium for Blood Pressure (ICBP) and the
UK Biobank (UKB), based on 757,601 individuals of Euro-
pean ancestry.
19
In the pooled sample, mean systolic BP
(SBP) and diastolic BP (DBP) were 138.4 (SD 21.5) and 82.8
(SD 11.4) mm Hg, respectively. As genetic instruments for
SBP and DBP, we selected single nucleotide polymorphisms
(SNPs) associated with SBP or DBP at genome-wide signif-
icance level (p<5×10
−8
) and clumped for linkage disequi-
librium (LD) to r
2
< 0.001 based on the European 1,000
Genomes panel. We estimated the proportion of variance in
SBP and DBP explained by each instrument
22
and calculated
F statistics to measure instrument strength (tables e-1 and e-2,
doi.org/10.5061/dryad.dfn2z34wj).
23
We further selected genetic variants as proxies for the SBP-
lowering effects of common antihypertensive drug classes
(figure 1). According to our previously described strategy,
18
we identified the genes encoding pharmacologic targets re-
lated to BP-lowering for common antihypertensive drug
classes in DrugBank
24
and screened the genomic regions
corresponding to these genes and their regulatory regions
(promoters and enhancers).
25
For the main analyses, we se-
lected SNPs associated with SBP at genome-wide significance
(p<5×10
−8
) that were at moderate to low LD (r
2
< 0.4)
according to previously described approaches,
26–28
with
sensitivity analyses using a more stringent LD threshold (r
2
<
0.1) (table e-3, doi.org/10.5061/dryad.dfn2z34wj). The
genes and the specific genomic regions screened for identifi-
cation of genetic proxies for each antihypertensive drug class
are detailed in table e-4 (doi.org/10.5061/dryad.dfn2z34wj).
Primary outcomes and etiologically
related phenotypes
The primary outcomes for our analyses were any stroke, is-
chemic stroke and its Trial of Org 10172 in Acute Stroke
Treatment (TOAST)–defined subtypes (LAS, CES, SVS),
29
or ICH and its location-specific subtypes, i.e. lobar (origi-
nating at cerebral cortex or cortical–subcortical junction) and
deep (originating at thalamus, internal capsule, basal ganglia,
deep periventricular white matter, cerebellum, or brain-
stem).
30
Genetic association estimates for any stroke, ische-
mic stroke, and its subtypes were obtained from the
Glossary
ACE = angiotensin-converting enzyme; BB =β-blockers; BP = blood pressure; CCB = calcium channel blocker; CES =
cardioembolic stroke; DBP = diastolic blood pressure; GWAS = genome-wide association studies; ICBP = International
Consortium for Blood Pressure; ICH = intracerebral hemorrhage; ISGC = International Stroke Genetics Consortium; IVW =
inverse variance weight; LAS = large artery stroke; LD = linkage disequilibrium; MR-PRESSO = Mendelian randomization
pleiotropy residual sum and outlier; OR = odds ratio; RCT = randomized controlled trial; SBP = systolic blood pressure; SNP =
single nucleotide polymorphism; SVD = small vessel disease; SVS = small vessel stroke; UKB = UK Biobank; WMH = white
matter hyperintensity.
2Neurology | Volume 95, Number 4 | July 28, 2020 Neurology.org/N
MEGASTROKE multiethnic GWAS meta-analysis of 67,162
cases (60,341 ischemic stroke, 6,688 LAS, 9,006 CES, 11,710
SVS) and 454,450 controls.
20,31
For ICH, we used the sum-
mary statistics from the International Stroke Genetics Con-
sortium (ISGC) meta-analysis by Woo et al.
30
including 1,545
cases (664 lobar, 881 deep) and 1,481 controls. In addition,
we performed Mendelian randomization analyses for the ra-
diologic phenotype of WMH volume, a manifestation of ce-
rebral SVD etiologically related to SVS and ICH. We
performed a GWAS analysis for total volume of WMH, de-
rived from T1 and T2 fluid-attenuated inversion recovery
images in the UKB data following a previously described
approach,
32
as detailed in e-Methods (doi.org/10.5061/
dryad.dfn2z34wj).
Statistical analysis
For SBP and DBP, we calculated individual Mendelian ran-
domization estimates and standard errors from the SNP–
exposure and SNP–outcome associations using the Wald es-
timator and the Delta method; second-order weights were
used.
33
The Mendelian randomization associations for SBP
and DBP with the primary outcomes were estimated by
pooling individual Mendelian randomization estimates using
fixed-effects inverse variance weighted (IVW) meta-analy-
ses.
33
All Mendelian randomization associations between
SBP, DBP, and stroke were scaled to 10 mm Hg increment in
SBP and 5 mm Hg in DBP.
For the antihypertensive drug classes, including instruments
at moderate to low LD (r
2
< 0.4), we applied generalized
linear regression analyses weighted for the correlation be-
tween the instruments, as previously described.
26
This rel-
atively lenient LD correlation threshold allows for an
increase in proportion of variance explained and thus in
statistical power.
26,27
In sensitivity analyses, we restricted
our instrument selection to a lower LD correlation thresh-
old (r
2
< 0.1) and applied fixed-effects IVW. All Mendelian
randomization associations between antihypertensive drug
classes and stroke were scaled to 10 mm Hg decrease
in SBP.
Mendelian randomization analyses might be biased due to
pleiotropic instruments. As measures of pleiotropy, we
Table 1 Descriptive characteristics of the genome-wide association study (GWAS) meta-analyses that were included in
this Mendelian randomization study
Study stage GWAS Phenotype Sample size Ancestry Adjustments
a
Instrument selection ICBP and UK Biobank
19
SBP, DBP 757,601 individuals European Age, sex, BMI
Use of instruments for
sensitivity analysis
UK Biobank (Neale
laboratory analysis)
39
SBP, DBP 317,756 individuals European None
Primary outcome MEGASTROKE
20
Any stroke, IS, and subtypes
(LAS, CES, SVS)
67,162 cases/
454,450 controls
Multiancestry/
European
Age, sex
Primary outcome ISGC ICH GWAS
30
ICH and subtypes (lobar,
deep ICH)
1,545 cases/1,481
controls
European Age, sex
Etiologically related
outcome
UK Biobank WMH volume 10,597 individuals European Age, sex
Abbreviations: BMI = body mass index; CES = cardioembolic stroke; DBP = diastolic blood pressure; ICBP = International Consortium for Blood Pressure;ICH=
intracerebral hemorrhage; IS = ischemic stroke; ISGC = International Stroke Genetics Consortium; LAS = large artery stroke; SBP = systol ic blood pressure; SVS
= small vessel stroke; WMH = white matter hyperintensities.
a
All GWAS studies have further adjusted for principal components.
Figure 1 Selection strategy for genetic variants used as
proxies for antihypertensive drug classes
Steps for genetic instrument selection and the respective criteria and
resources. ACC = American College of Cardiology; AHA = American Heart
Association; CES = cardioembolic stroke; ESC = European Society of Cardi-
ology; ESH = European Society of Hypertension; GWAS = genome-wide as-
sociation studies; ICBP = International Consortium for Blood Pressure; ICH =
intracerebral hemorrhage; LAS = large artery stroke; MR = Mendelain ran-
domization; SBP = systolic blood pressure; SVS = small vessel stroke; UKB =
UK Biobank; WMH = white matter hyperintensity.
Neurology.org/N Neurology | Volume 95, Number 4 | July 28, 2020 3
assessed heterogeneity across Mendelian randomization
estimates with I
2
and the Cochran Q test (I
2
> 50% and p<
0.05 were considered statistically significant)
34
and the in-
tercept obtained from Mendelian randomization–Egger re-
gression (p< 0.05 considered statistically significant).
35
We
further used alternative methods (weighted–median esti-
mator,
36
Mendelian randomization–Egger,
35
weighted–
modal estimator
37
) with relaxing assumptions regarding
pleiotropic variants. The weighted median estimator
requires that at least half of the information for the Men-
delian randomization analysis comes from valid instru-
ments.
36
Mendelian randomization–Egger regression
requires that the strengths of potential pleiotropic instru-
ments are independent of their direct associations with the
outcome.
35
The weighted modal estimator provides correct
estimates under the assumption that a plurality of genetic
variants are valid instruments.
37
We further tested for the
presence of pleiotropic outlier variants using the Mendelian
randomization pleiotropy residual sum and outlier (MR-
PRESSO) test
38
and in sensitivity IVW Mendelian ran-
domization analyses excluded these variants.
The genetic association estimates used in the analyses for
BP were corrected for antihypertensive medication use and
were adjusted for body mass index,
19
thus introducing po-
tential bias due to medication noncompliance or collider
effects, respectively. Thus we performed sensitivity analyses
using unadjusted estimates for BP from a UKB GWAS
(317,756 individuals).
39
To minimize ancestral mismatch
with the European population used in the BP GWAS, in
sensitivity analyses we further restricted our Mendelian
randomization analyses for stroke to the MEGASTROKE
European subset.
Statistical significance for all analyses was set at a 2-sided p
value <0.05. To examine whether BP differentially associated
with stroke subtypes or whether there were differential effects
of antihypertensive drugs on stroke risk, we compared the
derived odds ratios (ORs) by computing zscore for the dif-
ferences of their natural logarithms. All statistical analyses
were undertaken in R (v3.5.0; The R Foundation for Statis-
tical Computing) using the MendelianRandomization, Two-
SampleMendelian randomization, and MR-PRESSO
packages.
Data availability
This study was based on summary statistics. The GWAS data
from the ICBP and UKB meta-analysis are publicly available
through the GRASP repository of the National Heart, Lung,
and Blood Institute (grasp.nhlbi.nih.gov/FullResults.aspx).
The data from the GWAS studies for stroke and ICH are
publicly available and may be accessed through the MEGA-
STROKE (megastroke.org/download.html) and the ISGC
(cerebrovascularportal.org/informational/downloads) web-
sites, respectively. Data from the UKB GWAS for WMH
volume may be accessed through an application to UKB. The
summary data for the genetic instruments used for the
purposes of the current study are available in tables e-1 to e-3
(doi.org/10.5061/dryad.dfn2z34wj).
Results
Genetically determined BP and risk of
stroke subtypes
We first examined the relationship between genetically de-
termined BP and the risk of stroke and stroke subtypes. We
identified 462 genetic variants associated with SBP and 460
variants associated with DBP. F statistic was >10 for all
variants, indicating low risk of weak instrument bias (tables
e-1 and e-2, doi.org/10.5061/dryad.dfn2z34wj). Mendelian
randomization analyses showed statistically significant
associations of both SBP and DBP with risk of any stroke,
ischemic stroke, and all of its major subtypes (LAS, CES,
SVS), ICH, and deep ICH, but not lobar ICH (figure 2). The
effects of genetically determined BP were larger for LAS and
SVScomparedtoCES(pfor LAS-CES comparisons of ORs
=2×10
−8
for SBP and 0.004 for DBP; pfor SVS-CES
comparisons of ORs = 0.001 for SBP and 9 × 10
−4
for DBP),
and for deep compared to lobar ICH (pfor comparisons of
ORs = 0.016 for SBP and 0.009 for DBP), as depicted in
figure 2.
The effect estimates remained stable in the weighted median,
MR-Egger, and weighted-modal analyses, analyses excluding
outliers detected with MR-PRESSO, European-restricted
analyses, and analyses based on unadjusted BP estimates
(table e-5, doi.org/10.5061/dryad.dfn2z34wj). Tests for
heterogeneity and the MR-Egger intercepts were not signifi-
cant in any of the analyses (I
2
< 50% and p> 0.05, re-
spectively), providing no evidence for pleiotropy.
Genetic proxies for antihypertensive drugs and
risk of stroke subtypes
Next, we selected BP-lowering variants in genes encoding
drug targets as proxies for the effects of antihypertensive drug
classes, as detailed in figure 1 and as has been previously
described,
18
and examined their effects on stroke in Mende-
lian randomization analyses. We identified 8 proxies (var-
iants) for BBs and 60 proxies for CCBs (table e-3, doi.org/10.
5061/dryad.dfn2z34wj). We further identified a single proxy
for angiotensin-converting enzyme (ACE) inhibitors, which
we did not consider in the following analyses given the lack of
power. A 10-mm Hg reduction in SBP through variants in
genes encoding targets of CCBs, but not BBs, was associated
with a significantly lower risk of any stroke and ischemic
stroke (figure 3). In analyses for ischemic stroke subtypes, we
found a 10-mm Hg reduction in SBP through CCB variants to
be associated with significantly lower risks of LAS, CES, and
SVS. The effect for SVS was stronger than that for both LAS
(pfor comparison of ORs = 0.002) and CES (pfor com-
parison of ORs = 6 × 10
−4
)(figure 3). BB variants were not
associated with any of the ischemic stroke subtypes. We found
no significant associations for any of the drug classes for ICH
4Neurology | Volume 95, Number 4 | July 28, 2020 Neurology.org/N
and its subtypes, which is probably related to limited power
(table e-6, doi.org/10.5061/dryad.dfn2z34wj).
Sensitivity analyses for BBs and CCBs restricted to the set of
variants with a more stringent LD threshold (r
2
< 0.1) showed
consistent association estimates with the primary analyses for
all of the examined phenotypes (table e-6, doi.org/10.5061/
dryad.dfn2z34wj). For CCBs, we found no evidence for
pleiotropy (heterogeneity: I
2
< 50%; pof MR-Egger inter-
cepts > 0.05). There was heterogeneity in the associations of
BBs with any stroke (I
2
= 59%), ischemic stroke (I
2
= 67%),
and SVS (I
2
= 66%), which was however attenuated following
exclusion of 2 outlier SNPs in MR-PRESSO (I
2
= 0%, fol-
lowing exclusion of outlier SNPs), while the association
estimates remained stable (table e-6, doi.org/10.5061/dryad.
dfn2z34wj). The results remained consistent across the al-
ternative Mendelian randomization methods (table e-6, doi.
org/10.5061/dryad.dfn2z34wj).
Genetically determined BP and WMH volume
To gain additional insight into the relationship between
genetically determined BP and cerebral SVD, we next cal-
culated Mendelian randomization estimates for the associ-
ations of BP with WMH volume. We found genetically
elevated SBP and DBP to be significantly associated with
higher WMH volume (figure 4A). Examining the effects of
genetic proxies for antihypertensive drug classes (figure 4B),
we found significant associations of CCBs with lower WMH
volume (β=−0.491, 95% confidence interval −0.591 to
−0.391, p= 3.5 × 10
−7
), whereas proxies for BBs were not
associated with WMH volume. The results were consistent
across sensitivity analyses (table e-5, doi.org/10.5061/
dryad.dfn2z34wj).
Discussion
We investigated the relationship between the leading modi-
fiable risk factor for stroke and etiologically defined stroke
subtypes by leveraging large-scale genetic data. We found
Figure 2 Mendelian randomization associations between genetically determined blood pressure and risk of stroke and
stroke subtypes
Results from the fixed-effects inverse variance weighted analysis. CI = confidence interval; DBP = diastolic blood pressure; OR = odds ratio; SBP = systolic blood
pressure.
Figure 3 Mendelian randomization associations between
genetic proxies for antihypertensive drug classes
and risk of stroke and stroke subtypes
Results from the Mendelian randomization analysis adjusting for correla-
tion between variants. BB = β-blockers; CCB = calcium channel blocker; CI =
confidence interval; OR = odds ratio; SBP = systolic blood pressure.
Neurology.org/N Neurology | Volume 95, Number 4 | July 28, 2020 5
genetic predisposition to higher BP to be associated with
greater risk of any stroke, ischemic stroke, each of its main
subtypes, and deep but not lobar ICH. Risk was higher for
LAS and SVS compared to CES. Using genetic proxies for
different antihypertensive drug classes, we found BP-lowering
through CCBs, but not BBs, to be associated with lower risk of
stroke and ischemic stroke. CCB variants were associated with
a lower risk of all major ischemic stroke subtypes, showing
particularly strong effects on SVS and the related phenotype
of WMH.
Our study provides evidence for a causal effect of higher BP on
LAS, CES, and SVS, thus demonstrating a broad involvement
of BP in the pathogenesis of ischemic stroke. Of note, however,
we found the effects on stroke risk to vary depending on stroke
mechanisms. Specifically, risk was more pronounced for LAS
and SVS than for CES and was restricted to deep ICH. Unlike
deep ICH, lobar ICH is often related to cerebral amyloid
angiopathy and the absence of an association signal between BP
and lobar ICH is consistent with observational data.
40,41
As
demonstrated by our drug target analyses, the effects of specific
antihypertensive drug classes also differed according to stroke
subtype. Collectively, these data emphasize the need to con-
sider stroke etiologies when studying the effects of BP on stroke
risk in observational and interventional studies.
Among the major findings is a benefit of BP lowering through
genetic proxies for CCBs over BBs for SVS and the related
phenotype of WMH. In contrast, we found no disparity in
effects between genetic proxies for CCBs and BBs for LAS and
CES. This suggests that CCBs may be particularly effective in
preventing manifestations of cerebral SVD. The mechanisms
underlying this observation are unknown but may include
direct effects of CCBs on cerebral microvessels or systemic
effects, for instance, from the established influence of CCBs
on BP variability.
9,10,42
Patients with cerebral SVD mark a population at increased
risk for stroke, dementia, and death.
43
SVD manifestations are
highly prevalent in the aging population, with figures reaching
up to 90% in patients aged 65 years and above.
44
Yet there
have been no informative trials on specific antihypertensive
agents for the prevention of SVS, WMH, or other manifes-
tations of SVD.
45–47
Our Mendelian randomization results
suggest that BP lowering with CCBs should be tested in
clinical trials for prevention of SVS and other outcomes re-
lated to SVD.
The consistency of our results for stroke obtained from ge-
netic proxies for different drug classes with those from pre-
vious RCTs
7,9,10
is worth noting and lends confidence to our
findings on etiologic stroke subtypes for which no data from
RCTs exist. The disparity in treatment effects between CCBs
and BBs on stroke risk has been related to the opposite actions
of these drugs on BP variability; CCBs decrease whereas BBs
increase BP variability.
9,10
However, whether the effects of BP
variability on stroke risk vary by stroke etiology is unresolved
and deserves further investigation.
Our study has several methodologic strengths. We used large
datasets offering sufficient statistical power for most analyses
and applied multiple methods to exclude pleiotropic effects
and other biases. We also examined phenotypes etiologically
related to stroke subtypes and performed mediation analyses
Figure 4 Mendelian randomization associations of (A) genetically determined blood pressure and (B) genetic proxies for
antihypertensive drug classes with WMH volume
Results from (A) the fixed effects inverse variance weighted
analysis and (B) Mendelian randomization analysis adjust-
ing for correlation between variants. BB = beta blockers;
CCB = calcium channel blockers; CI = confidence interval;
DBP = diastolic blood pressure; SBP = systolic blood
pressure.
6Neurology | Volume 95, Number 4 | July 28, 2020 Neurology.org/N
that allowed inferences on mechanistic aspects regarding the
association of BP with stroke. Finally, we used genetic proxies
for antihypertensive drug classes that have been validated
previously and have shown comparable effects to those de-
rived from RCTs.
18
Our study also has limitations. First, Mendelian randomization
examines the lifetime effects of genetically determined BP,
which might differ from the effect of a clinical intervention for
BP lowering. Second, based on our selection criteria, we
identified only a single genetic proxy for ACE inhibitors that
did not offer sufficient statistical power to perform meaningful
analyses. Future studies encompassing larger GWAS datasets
for BP might identify such variantsand might thus offer deeper
insights into differential effects between different classes of BP-
lowering agents including ACE inhibitors, angiotensin-receptor
blockers, and thiazide diuretics on stroke and stroke subtypes.
Third, by design, we could not examine nonlinear associations
between BP and stroke risk.
48
However, current evidence
suggests that the association of midlife SBP and DBP with
stroke seems to follow a linear pattern.
49
Fourth, our results
apply stroke incidence and not stroke recurrence. While we
found high BP to not be associated with risk of lobar ICH,
hypertension has been shown in observational studies to in-
crease the risk for both deep and lobar ICH recurrence,
50
which
could not be examined in the context of the current study. Fifth,
the small sample size for the ICH GWAS did not offer sufficient
power to examine the effects of antihypertensive drug classes
on any, lobar, and deep ICH. Sixth, our GWAS data for BP
were restricted to individuals of European ancestry, which
could limit generalizability of our findings to this population.
This might specifically apply for ICH
30
given the evidence from
observational studies for differential associations of BP with
lobar ICH depending on ethnicity.
51
Furthermore, there is
evidence for differential responses to antihypertensive drug
classes by ethnicity, which could not be examined in the current
study.
52
The availability of large-scale GWAS data from more
diverse populations with higher representation of non-
European ethnicities will enable future Mendelian randomiza-
tion studies to explore potential ethnic disparities in more
detail. Finally, it was not possible to disentangle the effects of
dihydropyridine and nondihydropyridine CCBs with Mende-
lian randomization, because the differences in the subunits of
the voltage-gated calcium channels that are the targets of these
drug subclasses in the vessels and the heart, respectively, are
encoded by the same genes but are the result of alternative
splicing.
53
We provide evidence for a causal association of higher BP
with risk of any stroke and all stroke subtypes except lobar
ICH, with a higher risk of large artery stroke and SVS
compared to cardioembolic stroke. Our findings support
CCBs, but not BBs, to lower ischemic stroke risk. Genetic
proxies for the effects of CCBs showed particularly strong
associations with SVS and WMH, highlighting calcium
channel blockade as a promising strategy for the prevention
of cerebral SVD.
Acknowledgment
This research has been conducted using the UK Biobank
resource (UK Biobank application 2532). The authors thank
the contributions by the MEGASTROKE Consortium, Woo
et al. for the ICH GWAS meta-analysis, the ICBP
Consortium, the CHARGE Consortium, the AFGen Con-
sortium, and the Neale laboratory for performing GWAS
analyses in the UK Biobank data. MEGASTROKE has
received funding from the sources, detailed at megastroke.
org/acknowledgments.html. All MEGASTROKE authors and
their affiliations are available at megastroke.org/authors.html.
Study funding
M. Georgakis is funded by scholarships from the German
Academic Exchange Service (DAAD) and Onassis Founda-
tion. D. Gill is funded by the Wellcome Trust. P. Elliott
acknowledges support from the Medical Research Council
and Public Health England (Mendelian randomization/
L01341X/1) for the C-PHE Centre for Environment and
Health. P. Elliott is supported by the UK Dementia Research
Institute, which receives its funding from UK DRI Ltd. funded
by the UK Medical Research Council, Alzheimer’s Society,
and Alzheimer’s Research UK; and the National Institute of
Health Research Imperial College Biomedical Research
Centre. P. Elliott is associate director of the Health Data
Research UK London funded by a consortium led by the UK
Medical Research Council. This project has received funding
from the European Union’s Horizon 2020 research and in-
novation programme (no. 666881), SVDs@target (to M.
Dichgans), and no. 667375, CoSTREAM (to M. Dichgans);
the DFG as part of the Munich Cluster for Systems Neurology
(EXC 1010 SyNergy) and the CRC 1123 (B3) (to M.
Dichgans); the Corona Foundation (to M. Dichgans); the
Fondation Leducq (Transatlantic Network of Excellence on
the Pathogenesis of Small Vessel Disease of the Brain) (to M.
Dichgans); the e:Med program (e:AtheroSysMed) (to M.
Dichgans); and the FP7/2007–2103 European Union project
CVgenes@target (grant agreement number Health-F2-2013-
601456) (to M. Dichgans).
Disclosure
The authors report no disclosures relevant to the manuscript.
Go to Neurology.org/N for full disclosures.
Publication history
Received by Neurology June 28, 2019. Accepted in final form
January 5, 2020.
Appendix Authors
Name Location Contribution
Marios K.
Georgakis,
MD
LMU Munich,
Germany
Concept and design; data acquisition,
analysis, and interpretation; statistical
analysis; drafting of the manuscript;
critical revision of the manuscript for
intellectual content
Continued
Neurology.org/N Neurology | Volume 95, Number 4 | July 28, 2020 7
References
1. GBD 2015 DALYs and Hale Collaborators. Global, regional, and national disability-
adjusted life-years (DALYs) for 315 diseases and injuries and healthy life expectancy
(HALE), 1990-2015: a systematic analysis for the Global Burden of Disease Study
2015. Lancet 2016;388:1603–1658.
2. GBD 2015 Mortality and Causes of Death Collaborators. Global, regional, and na-
tional life expectancy, all-cause mortality, and cause-specific mortality for 249 causes
of death, 1980-2015: a systematic analysis for the Global Burden of Disease Study
2015. Lancet 2016;388:1459–1544.
3. O’Donnell MJ, Chin SL, Rangarajan S, et al. Global and regional effects of potentially
modifiable risk factors associated with acute stroke in 32 countries (INTER-
STROKE): a case-control study. Lancet 2016;388:761–775.
4. Feigin VL, Roth GA, Naghavi M, et al. Global burden of stroke and risk factors in 188
countries, during 1990-2013: a systematic analysis for the Global Burden of Disease
Study 2013. Lancet Neurol 2016;15:913–924.
5. Lewington S, Clarke R, Qizilbash N, Peto R, Collins R, Prospective Studies Collab-
oration. Age-specific relevance of usual blood pressure to vascular mortality: a meta-
analysis of individual data for one million adults in 61 prospective studies. Lancet
2002;360:1903–1913.
6. Lacey B, Lewington S, Clarke R, et al. Age-specific association between blood pressure
and vascular and non-vascular chronic diseases in 0.5 million adults in China: a pro-
spective cohort study. Lancet Glob Health 2018;6:e641–e649.
7. Ettehad D, Emdin CA, Kiran A, et al. Blood pressure lowering for prevention of
cardiovascular disease and death: a systematic review and meta-analysis. Lancet 2016;
387:957–967.
8. Xie X, Atkins E, Lv J, et al. Effects of intensive blood pressure lowering on cardio-
vascular and renal outcomes: updated systematic review and meta-analysis. Lancet
2016;387:435–443.
9. Rothwell PM, Howard SC, Dolan E, et al. Effects of beta blockers and calcium-channel
blockers on within-individual variability in blood pressure and risk of stroke. Lancet
Neurol 2010;9:469–480.
10. Webb AJ, Fischer U, Mehta Z, Rothwell PM. Effects of antihypertensive-drug class on
interindividual variation in blood pressure and risk of stroke: a systematic review and
meta-analysis. Lancet 2010;375:906–915.
11. Zia E, Hedblad B, Pessah-Rasmussen H, Berglund G, Janzon L, Engstrom G. Blood
pressure in relation to the incidence of cerebral infarction and intracerebral hemor-
rhage. Hypertensive hemorrhage: debated nomenclature is still relevant. Stroke 2007;
38:2681–2685.
12. Ohira T, Shahar E, Chambless LE, Rosamond WD, Mosley TH Jr, Folsom AR. Risk
factors for ischemic stroke subtypes: the Atherosclerosis Risk in Communities study.
Stroke 2006;37:2493–2498.
13. Holmes MV, Ala-Korpela M, Smith GD. Mendelian randomization in car-
diometabolic disease: challenges in evaluating causality. Nat Rev Cardiol 2017;14:
577–590.
14. Roberts R. Mendelian randomization studies promise to shorten the journey to FDA
approval. JACC Basic Transl Sci 2018;3:690–703.
15. Georgakis MK, Gill D, Rannikmae K, et al. Genetically determined levels of circulating
cytokines and risk of stroke. Circulation 2019;139:256–268.
16. Ference BA, Robinson JG, Brook RD, et al. Variation in PCSK9 and HMGCR and risk
of cardiovascular disease and diabetes. N Engl J Med 2016;375:2144–2153.
17. Ference BA, Kastelein JJP, Ginsberg HN, et al. Association of genetic variants related
to CETP inhibitors and statins with lipoprotein levels and cardiovascular risk. JAMA
2017;318:947–956.
18. Gill D, Georgakis MK, Koskeridis F, et al. Use of genetic variants related to antihy-
pertensive drugs to inform on efficacy and side effects. Circulation 2019;140:270–279.
19. Evangelou E, Warren HR, Mosen-Ansorena D, et al. Genetic analysis of over 1 million
people identifies 535 new loci associated with blood pressure traits. Nat Genet 2018;
50:1412–1425.
20. Malik R, Chauhan G, Traylor M, et al. Multiancestry genome-wide association study
of 520,000 subjects identifies 32 loci associated with stroke and stroke subtypes. Nat
Genet 2018;50:524–537.
21. Davey Smith G, Davies NM, Dimou N, et al. STROBE-MR: guidelines for
strengthening the reporting of Mendelian randomization studies. PeerJ Preprints
2019;7:e27857v27851.
22. Shim H, Chasman DI, Smith JD, et al. A multivariate genome-wide association
analysis of 10 LDL subfractions, and their response to statin treatment, in 1868
Caucasians. PLoS One 2015;10:e0120758.
23. Palmer TM, Lawlor DA, Harbord RM, et al. Using multiple genetic variants as
instrumental variables for modifiable risk factors. Stat Methods Med Res 2012;21:
223–242.
24. Wishart DS, Feunang YD, Guo AC, et al. DrugBank 5.0: a major update to the
DrugBank database for 2018. Nucleic Acids Res 2018;46:D1074–D1082.
25. Fishilevich S, Nudel R, Rappaport N, et al. GeneHancer: Genome-wide Integration of
Enhancers and Target Genes in GeneCards. 2017. Oxford: Database; 2017.
26. Burgess S, Dudbridge F, Thompson SG. Combining information on multiple in-
strumental variables in Mendelian randomization: comparison of allele score and
summarized data methods. Stat Med 2016;35:1880–1906.
27. Burgess S, Ference BA, Staley JR, et al. Association of LPA variants with risk of
coronary disease and the implications for lipoprotein(a)-lowering therapies: a Men-
delian randomization analysis. JAMA Cardiol 2018;3:619–627.
28. Nowak C, Arnlov J. A Mendelian randomization study of the effects of blood lipids on
breast cancer risk. Nat Commun 2018;9:3957.
29. Adams HP Jr, Bendixen BH, Kappelle LJ, et al. Classification of subtype of acute
ischemic stroke: definitions for use in a multicenter clinical trial: TOAST: Trial of Org
10172 in Acute Stroke Treatment. Stroke 1993;24:35–41.
30. Woo D, Falcone GJ, Devan WJ, et al. Meta-analysis of genome-wide association
studies identifies 1q22 as a susceptibility locus for intracerebral hemorrhage. Am J
Hum Genet 2014;94:511–521.
31. Malik R, Rannikmae K, Traylor M, et al. Genome-wide meta-analysis identifies 3 novel
loci associated with stroke. Ann Neurol 2018;84:934–939.
32. Rutten-Jacobs LCA, Tozer DJ, Duering M, et al. Genetic study of white matter
integrity in UK Biobank (N=8448) and the overlap with stroke, depression, and
dementia. Stroke 2018;49:1340–1347.
33. Burgess S, Butterworth A, Thompson SG. Mendelian randomization analysis with
multiple genetic variants using summarized data. Genet Epidemiol 2013;37:658 –665.
34. Del Greco MF, Minelli C, Sheehan NA, Thompson JR. Detecting pleiotropy in
Mendelian randomisation studies with summary data and a continuous outcome. Stat
Med 2015;34:2926–2940.
35. Bowden J, Davey Smith G, Burgess S. Mendelian randomization with invalid
instruments: effect estimation and bias detection through Egger regression. Int J
Epidemiol 2015;44:512–525.
36. Bowden J, Davey Smith G, Haycock PC, Burgess S. Consistent estimation in Men-
delian randomization with some invalid instruments using a weighted median esti-
mator. Genet Epidemiol 2016;40:304–314.
37. Hartwig FP, Davey Smith G, Bowden J. Robust inference in summary data Mendelian
randomization via the zero modal pleiotropy assumption. Int J Epidemiol 2017;46:
1985–1998.
38. Verbanck M, Chen CY, Neale B, Do R. Detection of widespread horizontal pleiotropy
in causal relationships inferred from Mendelian randomization between complex
traits and diseases. Nat Genet 2018;50:693–698.
39. Flossmann E, Schulz UG, Rothwell PM. Systematic review of methods and results of
studies of the genetic epidemiology of ischemic stroke. Stroke 2004;35:212–227.
40. Jackson CA, Sudlow CL. Is hypertension a more frequent risk factor for deep than for
lobar supratentorial intracerebral haemorrhage?. J Neurol Neurosurg Psychiatry 2006;
77:1244–1252.
41. Martini SR, Flaherty ML, Brown WM, et al. Risk factors for intracerebral hemorrhage
differ according to hemorrhage location. Neurology 2012;79:2275–2282.
Appendix (continued)
Name Location Contribution
Dipender
Gill, MD
Imperial
College
London, UK
Concept and design; data acquisition,
analysis, and interpretation; statistical
analysis; critical revision of the
manuscript for intellectual content
Alastair J.S.
Webb, DPhil
University of
Oxford, UK
Data acquisition, analysis, and
interpretation; critical revision of the
manuscript for intellectual content
Evangelos
Evangelou,
PhD
University of
Ioannina,
Greece
Data acquisition, analysis, and
interpretation; critical revision of the
manuscript for intellectual content
Paul Elliott,
PhD
Imperial
College
London, UK
Data acquisition, analysis, and
interpretation; critical revision of the
manuscript for intellectual content
Cathie L.M.
Sudlow,
DPhil
University of
Edinburgh, UK
Data acquisition, analysis, and
interpretation; critical revision of the
manuscript for intellectual content
Abbas
Dehghan,
MD
Imperial
College
London, UK
Data acquisition, analysis, and
interpretation; critical revision of the
manuscript for intellectual content
Rainer Malik,
PhD
LMU Munich,
Germany
Concept and design; data acquisition,
analysis, and interpretation; statistical
analysis; critical revision of the
manuscript for intellectual content
Ioanna
Tzoulaki,
PhD
Imperial
College
London, UK
Concept and design; data acquisition,
analysis, and interpretation; statistical
analysis; critical revision of the
manuscript for intellectual content
Martin
Dichgans,
MD
LMU Munich,
Germany
Concept and design; data
acquisition, analysis, and
interpretation; drafting of the
manuscript; critical revision of the
manuscript for intellectual content
8Neurology | Volume 95, Number 4 | July 28, 2020 Neurology.org/N
42. Yamaguchi Y, Wada M, Sato H, et al. Impact of ambulatory blood pressure variability
on cerebral small vessel disease progression and cognitive decline in community-
based elderly Japanese. Am J Hypertens 2014;27:1257–1267.
43. Debette S, Schilling S, Duperron MG, Larsson SC, Markus HS. Clinical significance of
magnetic resonance imaging markers of vascular brain injury: a systematic review and
meta-analysis. JAMA Neurol 2019;76:81–94.
44. de Leeuw FE, de Groot JC, Achten E, et al. Prevalence of cerebral white matter lesions
in elderly people: a population based magnetic resonance imaging study: The Rot-
terdam Scan Study. J Neurol Neurosurg Psychiatry 2001;70:9–14.
45. Group SPSS, Benavente OR, Coffey CS, et al. Blood-pressure targets in patients with
recent lacunar stroke: the SPS3 randomised trial. Lancet 2013;382:507–515.
46. Croall ID, Tozer DJ, Moynihan B, et al. Effect of standard vs intensive blood pressure
control on cerebral blood flow in small vessel disease: the PRESERVE randomized
clinical trial. JAMA Neurol 2018;75:720–727.
47. van Middelaar T, Argillander TE, Schreuder F, Deinum J, Richard E, Klijn CJM. Effect
of antihypertensive medication on cerebral small vessel disease: a systematic review
and meta-analysis. Stroke 2018;49:1531–1533.
48. Kannel WB, Wolf PA, McGee DL, Dawber TR, McNamara P, Castelli WP. Systolic
blood pressure, arterial rigidity, and risk of stroke: The Framingham study. JAMA
1981;245:1225–1229.
49. Bangalore S, Messerli FH, Wun CC, et al. J-curve revisited: an analysis of blood
pressure and cardiovascular events in the Treating to New Targets (TNT) Trial. Eur
Heart J 2010;31:2897–2908.
50. BiffiA, Anderson CD, Battey TW, et al. Association between blood pressure
control and risk of recurrent intracerebral hemorrhage. JAMA 2015;314:
904–912.
51. Sawyer RP, Sekar P, Osborne J, et al. Racial/ethnic variation of APOE alleles for lobar
intracerebral hemorrhage. Neurology 2018;91:e410–e420.
52. Preston RA, Materson BJ, Reda DJ, et al. Age-race subgroup compared with renin
profile as predictors of blood pressure response to antihypertensive therapy: De-
partment of Veterans Affairs Cooperative Study Group on Antihypertensive Agents.
JAMA 1998;280:1168–1172.
53. Hu Z, Liang MC, Soong TW. Alternative splicing of L-type CaV1.2 calcium channels:
implications in cardiovascular diseases. Genes 2017;8:E344.
Neurology.org/N Neurology | Volume 95, Number 4 | July 28, 2020 9
