WSO December 2023 – Issue 3 - 1249

Liu et al.
1249
Outcome data
The primary outcome was dependence or death after
ischemic stroke, defined as mRS (modified Rankin scale)
3-6. The secondary outcome was disability or death after
ischemic stroke defined as mRS 2-6.22 The mRS evaluates
global disability with scores ranging from 0 (no symptoms)
to 6 (death). A score of 2-5 indicates increasing levels of
disability.23 We obtained the summary-level data for functional
outcomes after ischemic stroke from the Genetics of
Ischemic Stroke Functional Outcome (GISCOME) network
meta-analysis, which contains 12 studies with 6021 stroke
patients from the USA, Europe, and Australia.24 In the primary
analysis, we used the summary data which was
adjusted for age, sex, blood pressure, principal components,
and baseline National Institutes of Health Stroke Scale
(NIHSS) score. In the repeated analysis, we used the data
without adjustment for baseline NIHSS to verify our primary
results.24 All individuals were of European ancestry.
Statistical analysis
We conducted primary MR analysis using a random-effects
inverse variance weighted (IVW) meta-analysis to explore
the associations between BP and antihypertensive medications
and functional outcomes after ischemic stroke.
IVW MR method assumes that all variants are valid instruments25
and comprises a meta-analysis of SNP-specific
Wald estimates



β
β
SNP
SNP
−
−
outcome
exposure



using an inverse of the corresponding variance.26 We
excluded the SNPs that had F statistics lower than 10
according to standard practice.27 All MR associations
between SBP, DBP, PP, and primary outcomes were scaled
to 10 mmHg increment in SBP, 5 mmHg in DBP, and 1
mmHg in PP.
We performed several sensitivity analyses. First, we
used Cochran's Q statistic in the IVW model to assess the
heterogeneity between variant-specific estimates. If possible,
we used the results of the weighted median method.
The weighted median approaches give more weight to
more precise instrumental variables and the estimate is consistent
even when up to 50% of the information comes from
invalid or weak instruments.28 Afterwards, we used
MR-Egger and MR-Pleiotropy Residual Sum and Outlier
methods to explore horizontal29 pleiotropy. Finally, we conducted
leave-one-SNP-out analysis, in which SNPs were
systematically removed, to assess if results were driven by
a single SNP. The association is considered significant after
correction
for
multiple
testing
for
three BP indexes
(p < 0.016 (0.05/3)) and five antihypertensive medications'
classes (p < 0.01 (0.05/5)). We calculated all effect estimates
as adjusted odds ratios (OR) with 95% confidence
intervals (CIs). The study design is shown in Figure 1.
To validate the primary results, we repeated the analyses
in the outcome database without adjustment for baseline
NIHSS and estimated the effects of antihypertensive drugs
on functional outcomes after ischemic stroke by selecting
SNPs that were at a more stringent LD threshold (r2 < 0.1).
To eliminate potential bias due to medication noncompliance
or collider effects, we performed sensitivity analyses
using unadjusted estimates for BP from a UK Biobank
GWAS (317,756 individuals).30 Moreover, we included a
bi-directional MR analysis to explore whether functional
outcomes after ischemic stroke affect BP. All analyses were
performed via Two Sample MR (version 0.5.6), Mendelian
randomization (version 0.5.1), and MRPRESSO (version
1.0) packages in R version 4.2.1.31
Results
Genetically determined BP and functional
outcomes after ischemic stroke
There were 368, 374 and 328 independent genetic variants
verified to be associated with SBP, DBP, and PP, respectively
(Supplementary Tables S1-S3). Figure 2 illustrates
the result of the primary univariable MR analysis. A 10
mmHg increase of genetically determined SBP was associated
with disability or death after ischemic stroke (OR 1.29,
95% CI: 1.05-1.59, p = 0.014). A 5 mmHg increase of DBP
was associated with disability or death after ischemic stroke
(OR 1.27, 95% CI: 1.07-1.51, p = 0.006). Interestingly, each
1 mmHg increase of PP was associated with disability or
death after ischemic stroke and dependence or death after
ischemic stroke (OR 1.05, 95% CI: 1.02-1.08, p = 0.002;
International Journal of Stroke, 18(10)
Figure 1. Flowchart depicting study design.
WSO December 2023 – Issue 3

Table of Contents for the Digital Edition of WSO December 2023 – Issue 3
