Cognitive Decline and Blood Pressure
Systolic Blood Pressure Variation and Mean Heart Rate
Is Associated With Cognitive Dysfunction in Patients With
High Cardiovascular Risk
Michael Böhm, Helmut Schumacher, Darryl Leong, Giuseppe Mancia, Thomas Unger,
Roland Schmieder, Florian Custodis, Hans-Christoph Diener, Ulrich Laufs, Eva Lonn,
Karen Sliwa, Koon Teo, Robert Fagard, Josep Redon, Peter Sleight, Craig Anderson,
Martin O’Donnell, Salim Yusuf
See Editorial Commentary, pp 505–506
Downloaded from http://hyper.ahajournals.org/ by guest on July 31, 2017
Abstract—Elevated systolic blood pressure (SBP) correlates to cognitive decline and incident dementia. The effects of heart
rate (HR), visit to visit HR variation, and visit to visit SBP variation are less well established. Patients without preexisting
cognitive dysfunction (N=24 593) were evaluated according to mean SBP, SBP visit to visit variation (coefficient of
variation [standard deviation/mean×100%], CV), mean HR, and visit to visit HR variation (HR-CV) in the Ongoing
Telmisartan Alone and in Combination with Ramipril Global Endpoint Trial and the Telmisartan Randomized Assessment
Study in ACE Intolerant Subjects with Cardiovascular Disease. Cognitive function was assessed with mini mental state
examination. Cognitive dysfunction (fall in mini mental state examination ≤24 points), important cognitive decline (drop
of ≥5 points), and cognitive deterioration (drop of >1 point per year or decline to <24 points) were assessed. SBP and
HR were measured over 10.7±2.2 (mean±SD) visits. Mean SBP, mean HR, and SBP-CV were associated with cognitive
decline, dysfunction, and deterioration (all P<0.01, unadjusted). After adjustment, only SBP-CV (P=0.0030) and mean
HR (P=0.0008) remained predictors for cognitive dysfunction (odds ratios [95% confidence intervals], 1.32 [1.10–1.58]
for 5th versus 1st quintile of SBP-CV and 1.40 [1.18–1.66] for 5th versus 1st quintile of mean HR). Similar effects were
observed for cognitive decline and deterioration. SBP-CV and mean HR showed additive effects. In conclusion, SBPCV and mean HR are independent predictors of cognitive decline and cognitive dysfunction in patients at high CV risk.
Clinical Trial Registration—URL: http://www.clinicaltrials.gov. Unique identifier: NCT 00153101. (Hypertension.
2015;65:651-661. DOI: 10.1161/HYPERTENSIONAHA.114.04568.) Online Data Supplement
•
Key Words: heart rate ◼ hypertension ◼ myocardial infarction ◼ stroke
C
artery disease,9 and heart failure10,11 and to induce endothelial
dysfunction,12 antiangiogenic mechanisms in hypercholesterolemia after vascular occlusions,13 and to enhance stroke size in
experimental models.14 In patients after a first stroke, elevated
HR >76 bpm was associated with cardiovascular mortality
and with a more pronounced cognitive decline after a second
stroke.15 Although the effects of noncontrolled SBP on cognitive decline have been studied,2 limited or no such information
is available for SBP visit to visit variation,16 mean HR, and HR
ardiovascular risk factors, particularly elevated systolic
blood pressure (SBP), have been linked to cognitive dysfunction and cognitive decline in cardiovascular disease1,2 potentially involving mechanisms, such as endothelial dysfunction,
microemboli, and oxidative stress,3,4 promoting cerebral atherosclerosis.5 The visit to visit variation of blood pressure has
been suggested to be linked to stroke and to cognitive decline.6,7
Recently, elevated resting heart rates (HR) were found to be
related to cardiovascular outcomes in hypertension,8 coronary
Received September 10, 2014; first decision September 26, 2014; revision accepted November 28, 2014.
From the Klinik für Innere Medizin III, Universitätsklinikum des Saarlandes, Homburg, Germany (M.B., F.C., U.L.); Boehringer Ingelheim, Pharma
GmbH & Co. KG, Ingelheim, Germany (H.S.); Population Health Research Institute, McMaster University, Hamilton, Ontario, Canada (D.L., E.L.,
K.T., S.Y.); Centro di Fisiologica Clinica e Ipertensione, Universita Milano-Bicocca, Istituto Auxologico, Milan, Italy (G.M.); CARIM-School for
Cardiovascular Diseases, Maastricht University, Maastricht, The Netherlands (T.U.); Department of Nephrology and Hypertension, Friedrich-Alexander
University, Erlangen, Germany (R.S.); Department of Neurology, University Hospital Essen, Essen, Germany (H.-C.D.); Hatter Institute for Cardiovascular
Research in Africa & IIDMM, Faculty of Health Sciences, University of Cape Town, South Africa (K.S.); Hypertension Unit, KU Leuven University,
Leuven, Belgium (R.F.); University of Valencia, Spain (J.R.); Department of Cardiovascular Medicine, John Radcliffe Hospital, Oxford, United Kingdom
(P.S.); The George Institute for Global Health, University of Sydney and Royal Prince Alfred Hospital, Sydney, NSW, Australia (C.A.); and HRB Clinical
Research Facility Galway, National University of Ireland, Galway, Geata an Eolais, University Road, Galway, Ireland (M.O’D.).
This paper was sent to David A. Calhoun, Guest editor, for review by expert referees, editorial decision, and final disposition.
The online-only Data Supplement is available with this article at http://hyper.ahajournals.org/lookup/suppl/doi:10.1161/HYPERTENSIONAHA.
114.04568/-/DC1.
Correspondence to Michael Böhm, Universitätsklinikum des Saarlandes, Klinik für Innere Medizin III, Kardiologie, Angiologie und Internistische
Intensivmedizin, Kirrberger Str. 1, D 66424 Homburg/Saar, Germany. E-mail [email protected]
© 2015 American Heart Association, Inc.
Hypertension is available at http://hyper.ahajournals.org
DOI: 10.1161/HYPERTENSIONAHA.114.04568
651
652 Hypertension March 2015
visit to visit variation. In this article, we analyze the relationship
between mean on-treatment SBP, SBP-variation, mean on-treatment HR, and HR-variation and cognitive decline (≥5 points
reduction in the mini mental state examination [MMSE]) and
cognitive dysfunction defined as a score of ≤24 points indicative
of clinically relevant cognitive impairment. Cognitive deterioration was defined as reduction of ≥1 point per year or decline to
≤24 points). The latter end point is sensitive to the absolute level
of MMSE and to changes over time although accounting for different durations of follow-up. We used the pooled data set of the
Ongoing Telmisartan Alone and in Combination With Ramipril
Global Endpoint Trial (ONTARGET) and the Telmisartan
Randomized Assessment Study in ACE Intolerant Subjects With
Cardiovascular Disease (TRANSCEND).17,18
Methods
Studied Patients
Downloaded from http://hyper.ahajournals.org/ by guest on July 31, 2017
The primary objective and protocol of the ONTARGET and
TRANSCEND trials have been described in detail previously.17,18 In
brief, patients age ≥55 years without symptomatic heart failure at entry
and with a history of coronary artery disease, peripheral artery disease,
prior TIA or stroke or diabetes mellitus complicated by organ damage
were eligible for inclusion in the studies. In ONTARGET patients known
to be tolerant to ACE inhibitors were randomly assigned to ramipril 10
mg qd, telmisartan 80 mg qd, or the combination thereof (double-dummy
design), whereas in TRANSCEND patients intolerant to ACE inhibitors
were randomized to telmisartan 80 mg qd or matching placebo. Study
medication was given on top of preexisting treatment used by the treating
physicians and investigators. The other medications were given according to best clinical practice in the individual study centers. Investigators
were advised to maintain preexisting blood pressure medications targeting blood pressures of <130/80 mm Hg for patients with chronic kidney
disease and 140/90 mm Hg in the other individuals. Patients with ontreatment blood pressure >160/100 mm Hg at entry were not eligible.17,18
Visits were scheduled at 6 weeks and 6 months after randomization and
every 6 months thereafter. In ONTARGET, 25.620 patients from 733
centers in 40 countries were randomized, whereas in TRANSCEND,
5926 patients from a subset of 630 centers were randomized. The
median follow-up period in both studies was 56 months. The protocol
procedures and main results of ONTARGET and TRANSCEND have
been published previously17,18 showing comparable outcomes in the composite of cardiovascular death, myocardial infarction, stroke, or hospitalization for heart failure (time to first event) in the 3 active treatment
arms of ONTARGET and a lower event rate in the telmisartan group of
TRANSCEND which, however, was not significant.
The primary objective of this analysis was to evaluate the association of
in-trial mean SBP, visit to visit SBP variation, mean HR, and visit to visit
HR variation with cognitive decline and incident cognitive dysfunction
in high risk cardiovascular patients. At each visit in both studies, resting
blood pressure and HR was measured in duplicate in sitting position after
3 minutes of rest using an automated validated device (Omron, model
HEM 757; Omron corporation, Kyoto, Japan). In-trial blood pressure and
HR were averaged using mean measurements at each individual visit.
Visit to visit variation of BP and HR was calculated as coefficient of variation (CV), that is, the ratio of standard deviation (SD) and mean (CV=SD/
mean×100%). Measurements from all visits (including baseline values)
before the final MMSE were included, and ≥3 visits with data were required for inclusion in the analysis. The studies were approved by the
local ethics committees in accordance with the Declaration of Helsinki.
Cognitive function was evaluated in all patients by MMSE. The score
ranges from 0 to 30 with lower scores indicative of greater degree of cognitive impairment. MMSE was done at baseline, after 2 years and at the
penultimate visit (usually between 3 and 5 years). The majority of patients
were evaluated at 3 to 5.5 years (88.5%), and in only 11.5% of patients,
the 2-year MMSE was the last evaluation. In line with a previous report,15
cognitive dysfunction was defined as MMSE ≤24 points at the last available visit (2 years or penultimate), and cognitive decline was defined as
≥5 points decrease of MMSE. The MMSE and its changes are related
to future dementia, although its accuracy in the diagnosis of dementia is
weak.19,20 It has been widely used to determine cognitive changes over
time and is a feasible instrument in large clinical trials.21–24 Sensitivity
analyses pointed out that changes between 3 and 4 points decline might
reliably reflect a significant cognitive decline.22,23 To overcome these
uncertainties, we present a ≥5 point MMSE decline as indication for a
significant cognitive decline. Similar results were obtained when a cutoff
of 2 points was evaluated. To account for different follow-up periods, in
the combined outcome of cognitive deterioration, a decline in ≥1 MMSE
points per year or a reduction to ≤24 points was also evaluated.
The different treatment arms of ONTARGET and TRANSCEND
did not show any differences in cognitive function at baseline.24 To
explore the role of mean SBP, SBP-CV, mean HR, and HR-CV on
cognitive function, data of enrolled patients without cognitive dysfunction at baseline (MMSE >24 points) and ≥1 available follow-up
score was included in this analysis.
31 546 patients were enrolled in ONTARGET or TRANSCEND.
In 4166 patients, there was preexisting significant cognitive dysfunction (MMSE ≤24) or no MMSE score at baseline was available. In
further 2646 patients, no follow-up MMSE was available; in an additional 42 patients, <3 visits were available for the calculation of SBP/
HR mean and variation, and in further 99 patients, data for important
covariates (those to be used in model building) were missing. Thus,
24 593 patients remained in the present analysis (Figure 1). On average, SBP and HR measurements were available from 10.7±2.2 visits
(range 3–13) over an observation period of 53.1±11.5 months.
Statistical Analysis
Patients were subdivided in quintiles based on their mean SBP,
SBP-CV, mean HR, and HR-CV, allowing cohort statistical
evaluations with adequate group sizes. Baseline characteristics
were presented for quintiles, continuous data as means±standard
deviations, and categorical data as percentages. Quintiles were
tested for differences using ANOVA for continuous data and the
χ2 test for categorical data. For each of the 4 parameters (SBP
mean, SBP-CV, HR mean, and HR-CV), rates of cognitive dysfunction, decline, and deterioration were determined by quintiles
Figure 1. Trial profile of the cognition analysis of Ongoing
Telmisartan Alone and in Combination With Ramipril Global
Endpoint Trial (ONTARGET) and Telmisartan Randomized
Assessment Study in ACE Intolerant Subjects With
Cardiovascular Disease (TRANSCEND). HR indicates heart rate;
MMSE, mini mental state examination; and SBP, systolic blood
pressure.
Böhm et al Blood Pressure, Heart Rate, and Cognitive Decline 653
and tested for differences using the χ2 test. In addition, we developed a multivariate logistic regression model that included all
4 parameters of interest and the MMSE value at baseline and,
in addition, all potential predictors of cognitive dysfunction,
decline, and deterioration, particularly those displayed in Table.
A stepwise selection procedure with P limits of 0.25 for entry
into the model and 0.15 for stay in the model was applied; it
was further checked whether the resulting optimal model was
Table. Odds Ratios for (A) Cognitive Dysfunction (MMSE ≤24), (B) Cognitive Decline (ΔMMSE ≥5), and
cognitive deterioration (MMSE ≤24/ΔMMSE ≥1 per year)
A: Cognitive
Dysfunction
Variable
MMSE at baseline, per score unit
DBP CV quintiles
B: Cognitive
Decline
C: Cognitive
Deterioration
OR
95% CI
OR
95% CI
OR
0.65
(0.63–0.67)
1.07
(1.03–1.11)
0.75
P=0.068
P=0.12
95% CI
(0.73–0.77)
P=0.0037
Downloaded from http://hyper.ahajournals.org/ by guest on July 31, 2017
2 vs 1
0.84
(0.71–1.00)
0.80
(0.65–0.98)
0.81
(0.69–0.94)
3 vs 1
0.81
(0.68–0.97)
0.83
(0.68–1.02)
0.74
(0.64–0.87)
4 vs 1
0.89
(0.75–1.06)
0.89
(0.72–1.09)
0.83
(0.71–0.97)
5 vs 1
0.96
(0.81–1.15)
0.97
(0.79–1.20)
0.86
(0.73–1.01)
Age, per year
1.05
(1.04–1.06)
1.06
(1.05–1.07)
1.06
(1.049–1.062)
BMI, per unit
0.99
(0.98–1.00)
0.97
(0.93–1.00)
1.09
(0.98–1.20)
eGFR, per 10 U
Female vs Male
Ethnicity
1.15
(1.03–1.29)
P=0.029
P<0.0001
P=0.0070
Asian vs White
0.93
(0.80–1.08)
0.83
(0.70–1.00)
0.92
(0.80–1.05)
Black vs White
1.39
(1.02–1.88)
1.80
(1.27–2.54)
1.41
(1.07–1.86)
Other vs White
1.16
(0.99–1.36)
1.24
(1.02–1.50)
1.16
(1.00–1.34)
Physical activity
P=0.099
P=0.016
Everyday vs mainly sedentary
0.87
(0.76–0.99)
0.83
(0.74–0.94)
5–6 times/week vs mainly sedentary
0.89
(0.71–1.11)
0.81
(0.66–0.98)
2–4 times/week vs mainly sedentary
1.02
(0.88–1.18)
0.95
(0.84–1.09)
≤Once/week vs mainly sedentary
0.89
(0.74–1.07)
0.86
(0.73–1.01)
Formal education
P<0.0001
P<0.0001
P<0.0001
College/University vs ≤8 y
0.42
(0.35–0.50)
0.42
(0.34–0.50)
0.46
(0.40–0.54)
Trade/Technical vs ≤8 y
0.56
(0.48–0.66)
0.59
(0.49–0.70)
0.57
(0.50–0.66)
9–12 yr vs ≤8 y
0.61
(0.54–0.69)
0.58
(0.50–0.67)
0.64
(0.58–0.72)
Alcohol consumption
0.81
(0.72–0.91)
0.74
(0.65–0.85)
0.78
(0.70–0.86)
1.09
(0.97–1.22)
History of stroke
1.22
(1.08–1.38)
1.25
(1.09–1.45)
1.20
(1.07–1.34)
New stroke
2.43
(1.96–3.00)
2.66
(2.11–3.35)
2.21
(1.82–2.70)
History of diabetes mellitus
1.29
(1.10–1.50)
1.35
(1.12–1.61)
1.28
(1.11–1.47)
1.17
(0,96-1.44)
1.41
(1.01–1.97)
0.85
(0.76–0.95)
0.90
(0.79–1.04)
0.86
(0.77–0.95)
0.84
(0.74–0.96)
History of hypertension
New diabetes mellitus
Doubling of creatinine
Concomitant medication
ASA
Beta blockers
Diuretics
1.09
(0.98–1.22)
Nitrates
1.12
(1.00–1.25)
1.20
(1.05–1.37)
1.13
(1.02–1.25)
Statins
0.92
(0.82–1.02)
0.90
(0.79–1.02)
0.92
(0.83–1.01)
Hypoglycemics
0.87
(0.74–1.02)
0.86
(0.71–1.05)
0.88
(0.76–1.02)
c-Statistic of complete model
0.784
0.709
0.739
c-Statistic of model without BP/HR indices
0.781
0.700
0.734
P=0.0001
P=0.0002
P<0.0001
Likelihood ratio test
Predictors resulting from multivariate logistic regression model selection (global P value given for predictors with >2 categories).
ASA indicates aspirin; BMI, body mass index; BP, blood pressure; CV, coeffefficient of variation; DBP, diastolic blood pressur; eGFR,
estimated glomerular filtration rate; MMSE, mini mental statement examination; and OR, odds ratio.
654 Hypertension March 2015
minimal in terms of Akaike’s Information Criterion. The likelihood ratio test was used to check whether the addition of BP
and HR parameters (as described above) significantly improve
the predictive quality of the models; c-statistics are given as
well. The association between SBP-CV and mean HR as continuous variables and cognitive outcomes was also analyzed
nonparametrically with restricted cubic splines, allowing for
potentially nonlinear relationship.25 Additionally, as a sensitivity
analysis, decline in MMSE was modeled as a continuous variable (reported in the online-only Data Supplement). Results are
presented as odds ratios (OR) with 95% confidence intervals. All
analyses were done using SAS version 9.2 (SAS Institute, NC).
Results
Correlation of Means, Standard Deviations, and
CVs
Downloaded from http://hyper.ahajournals.org/ by guest on July 31, 2017
The correlation between mean SBP and SBP SD was modest
(r=0.28), as was the correlation between mean HR and HR SD
(r=0.31). There was almost no correlation between mean SBP
and SBP coefficient of variation (CV; r<0.01) and between
mean H and HR-CV (r=0.05), and therefore, CV was chosen for further analysis instead of SD being less dependent
on the mean. There was almost no correlation between mean
SBP and mean HR (r=0.01), whereas the correlation between
SBP-CV and HR-CV was small (r=0.15).
Demographics and Clinical Characteristics
Demographic data and clinical characteristics are presented for
mean SBP, SBP-CV quintiles mean HR, and HR-CV quintiles
(Table S1 in the online-only Data Supplement). Table S1 also
contains a column for those 6953 patients who could not be
included in the analysis. Because of the big sample size, most
of the variables and characteristics were significantly different
between quintiles, but the relevance of these (partly small) differences is unsure. As expected, there were some differences with
cardiovascular medication, for example beta blockers were used
more often at low HR. There were also some differences between
the patients included in this analysis and the remaining patients
of ONTARGET and TRANSCEND. Some of these differences
were as a result of the selection of patients with MMSE >24 at
study entry; excluding patients with preexisting cognitive dysfunction mainly affected patients with older age, less advanced
formal education, less physical activity, black ethnicity, and
worse kidney function. In line with the higher age of the excluded
patients, means of SBP/HR and CVs were also slightly higher.
Proportions of Cognitive Dysfunction, Decline, and
Deterioration by SBP Mean and CV
Cognitive dysfunction was observed in 1857 patients (7.6%)
and cognitive decline in 1176 patients (4.8%); 1017 patients
(4.1%) fulfilled the criteria for both end points. Regarding
mean SBP (Figure 2A, left), incident cognitive dysfunction
increased from 6.0% in the lowest to 9.4% in the highest quintile (P<0.0001). Cognitive decline (middle) ranged from 4.0%
in the lowest to 5.7% in the highest quintile (P=0.0012). For
the quintiles of SBP-CV (Figure 2B), cognitive dysfunction
rates increased from 6.3% in the lowest to 10.2% in the highest quintile (P<0.0001). Cognitive decline rates ranged from
4.1% in the lowest to 6.4% in the highest quintile (P<0.0001;
Figure 2B). Similar results were obtained for cognitive deterioration (right; 7.4%–11.2% for mean SBP [P<0.0001] and
7.8%–12.1% for SBP-CV [P<0.0001], respectively).
Proportions of Cognitive Dysfunction, Decline, and
Deterioration by HR Mean and CV
Regarding mean HR, incident cognitive dysfunction (Figure 2C,
left) increased from 5.6% in the lowest to 9.1% in the highest
quintile (P<0.0001). Cognitive decline (Figure 2C, middle)
ranged from 3.4% in the lowest to 5.6% in the highest quintile (P<0.0001), with similar results for cognitive deterioration
(6.8%–11.0%, right). For the quintiles of HR-CV (Figure 2D),
cognitive dysfunction rates (left) ranged from 6.9% in the second (ie, lowest rate) to 8.5% in the highest quintile (P=0.027).
Cognitive decline (middle) rates were evaluated from 4.0% to
5.4% in the fifth quintile of HR-CV (P=0.0098; Figure 2D).
Cognitive deterioration showed similar results (right).
Multivariate Model With Adjustment for
Confounders
With respect to the incidence of cognitive dysfunction, SBP-CV
(P=0.0030) and mean HR (P=0.0008) remained to be significant
predictors after adjusting. The confounders, such as MMSE, at
baseline, DBP-CV, age, body mass index, estimated glomerular
filtration rate, sex, ethnicity, physical activity, formal education,
alcohol consumption, history of stroke and new stroke during
study conduct, history of diabetes mellitus, and new diabetes mellitus during study conduct, as well as concomitant medication with
aspirin, beta blockers, diuretics, nitrates, statins, and hypoglycemics are given in Table. Compared with the lowest SBP-CV quintile (reference), the OR for having cognitive dysfunction was 1.32
(95% CI, 1.10–1.58; P=0.0029) in the highest quintile (Figure 3A,
left). When comparing the highest mean HR quintile to the lowest,
the OR was 1.40 (1.18–1.66; P<0.0001; Figure 3B, left). Among
the other predictive variables, baseline MMSE (0.65 [0.63–0.67]
per unit), age (1.05 [1.04–1.06] per year), formal education (eg,
college/university versus <8 years education 0.42 [0.35–0.49]),
and incident stroke during the studies (2.42 [1.96–3.00]) and history of previous stroke before entry into the studies (1.22 [1.08–
1.38]) had the strongest associations with cognitive dysfunction.
The ORs for all confounders that remained in the model after the
selection procedure are presented in Table, A.
Mean HR (P=0.0002; Figure 3B) also was an independent
predictor for cognitive dysfunction (left), decline (middle),
and deterioration (right) after adjusting for the other parameters and all confounders. For mean HR, the upper 3 quintiles
had significantly higher odds compared with the referent quintile, but with no apparent differences among these 3 quintiles
were seen. Patients with higher MMSE baseline values had
a greater chance for decline (OR, 1.07 [1.03–1.11] per unit).
The ORs for all other factors in the model are presented in
Table, B. Similar results were obtained for cognitive deterioration (Table, C; Figure 3, right). The nonlinear relationship
between SBP-CV as a continuous variable and the risk of cognitive outcomes (as a result of modeling with cubic splines)
is shown in Figure 4A. A similar nonlinear relationship was
observed with mean HR as shown in Figure 4B.
Böhm et al Blood Pressure, Heart Rate, and Cognitive Decline 655
A Mean Systolic Blood Pressure
Q2:
Q3:
Q4:
Q5:
v
v
v
r r
Q1:
B Systolic Blood Pressure - CV
Q2:
Q3:
Q4:
Q5:
v
v
v
C Mean Heart Rate
Q2:
Q3:
Q4:
Q5:
v
v
v
r r
Q1:
D Heart Rate - CV
Q2:
Q3:
Q4:
Q5:
v
v
r r
Q1:
v
Downloaded from http://hyper.ahajournals.org/ by guest on July 31, 2017
r r
Q1:
Figure 2. Cognitive dysfunction (left), cognitive decline (middle), and cognitive deterioration (right) according to mean systolic blood
pressure (A), systolic blood pressure coefficient of variation (CV; B), mean heart rate (C), and heart rate coefficient of variation (D).
Interaction With Stroke
No apparent heterogeneity in the effects of SBP-CV and mean
HR on cognitive dysfunction, decline, and deterioration was
seen in patients with or without previous stroke (before or
during the studies [P values for test on interaction 0.61 for
SBP-CV, 0.35 for mean HR on cognitive dysfunction, 0.76 for
SBP-CV, 0.097 for mean HR for cognitive decline, and 0.62
for SBP-CV, 0.29 for mean HR for cognitive deterioration]).
656 Hypertension March 2015
Downloaded from http://hyper.ahajournals.org/ by guest on July 31, 2017
Figure 3. Development of cognitive dysfunction (left), cognitive decline (middle), and cognitive deterioration (right) according to systolic
blood pressure coefficient of variation (CV; A) and mean heart rate (B) quintiles. Odds ratios of systolic blood pressure CV and mean
heart rate quintiles were adjusted for mini mental state examination (MMSE) at baseline, diastolic blood pressure variation, age, body
mass index, estimated glomerular filtration rate (Modification of Diet in Renal Disease [MDRD]), sex, ethnicity, physical activity, formal
education, alcohol consumption, history of stroke and stroke during study conduct, history of diabetes mellitus and new diabetes mellitus
during study conduct, and concomitant medications with aspirin, beta blockers, diuretics, nitrates, statins, and hypoglycemics as
significant parameters, as selected in a stepwise selection procedure to identify the best multivariate logistic regression model.
Thus, the effects of high SBP variation and high mean HR
apply similarly to patients with and without stroke.
Interaction With Atrial Fibrillation
No apparent heterogeneity in the effects of SBP-CV and mean
HR on cognitive dysfunction and decline was seen in patients
with or without atrial fibrillation (before or during the studies
[P values for test on interaction 0.40 for SBP-CV, 0.43 for
mean HR on cognitive dysfunction, 0.31 for SBP-CV, 0.11
for mean HR for cognitive decline]). Thus, the effects of high
SBP variation and high mean HR apply similarly to patients
with and without atrial fibrillation. Only for the association
between mean HR and cognitive deterioration, there is some
indication (P=0.042) for a heterogeneous effect; the detrimental effect of higher mean HR seems to be confined to patients
without atrial fibrillation.
Interaction With Ethnicity
No heterogeneity in the effects of SBP-CV and mean HR was
seen in patients with different ethnicities on cognitive dysfunction (P -values for test on interaction 0.91 for SBP-CV, 0.22
for mean HR) and of SBP-CV on cognitive decline (P=0.44)
and cognitive deterioration (P=0.99). However, the effect of
mean HR on cognitive decline and deterioration seems to be
differential between ethnicities (P=0.025 for decline, P=0.022
for deterioration); the detrimental effect of high mean HR is
evident only in whites and Asians, whereas no effect is seen in
blacks and patients of other ethnic origin.
Interaction of Blood Pressure Variation and Mean
HR on Cognitive Function
To test whether there is any interaction between SBP-CV and
mean HR, the respective interaction term was added to the
multivariate model, which resulted from the model selection
process described above. It turned out that no apparent interaction was present (P=0.55 for new cognitive dysfunction,
P=0.11 for cognitive decline, P=0.70 for cognitive deterioration). The additivity in the detrimental effect of high SBP-CV
and high mean HR is shown in Figure 5, which shows ORs for
combination of quintiles with the combination of first quintiles as reference for cognitive dysfunction (A), decline (B),
and deterioration (C). Patients in the highest quintiles for each
parameter show an OR of 1.85 (1.44–2.38) for cognitive dysfunction; for cognitive decline, the maximum OR is seen in
Böhm et al Blood Pressure, Heart Rate, and Cognitive Decline 657
Downloaded from http://hyper.ahajournals.org/ by guest on July 31, 2017
Figure 4. Development of cognitive dysfunction (left), cognitive decline (middle), and cognitive deterioration (right) according to systolic
blood pressure coefficient of variation (SBP-CV; A) and mean heart rate (B) as continuous variables. Odds ratios calculated based on
logistic regression with restricted cubic splines (SBP-CV reference, 6%; mean heart rate reference, 60 bpm; gray area is 95% confidence
band).
the combined fifth SBP-CV and fourth HR mean quintile (2.03
[1.50–2.75]). The detailed results are given in Table S2A–2C.
Sensitivity Analyses
When the analyses described above are run in the subset of
patients with ≥3 visits subsequent the 6 week visit (n=24 365),
the results are essentially the same. This has been done to
reflect the possibility that SBP variation might be induced
by BP reduction in the beginning of the trials. Further, in a
linear model with (continuous) reduction in MMSE per year
as dependent variable and the same variables, which were
included in the logistic regression models described above,
Figure 5. Additive effects of systolic blood pressure variation (SBP-CV) and mean heart rate (HR mean) on cognitive dysfunction (A),
cognitive decline (B), and cognitive deterioration (C).
658 Hypertension March 2015
as independent variables, the significant effect of SBP-CV
(P<0.0001) and mean HR (P=0.0018) as well as the effects
of the main predictors (baseline MMSE, age, education, and
stroke) was confirmed (cf. Table S3). We also explored whether
our results were dependent on the definition of cut-offs, for
example, 25 points for MMSE at study end and 1 point for
yearly reduction. Figure S1 shows that the upper quintiles of
SBP-CV and mean HR show consistently higher cumulative
percentages of patients being <27 at final MMSE (Figure S1A
and 1C) and also higher cumulative percentages of patients
with yearly reductions in MMSE independent of its magnitude (Figure S1B and 1D). Thus, the effects demonstrated are
not dependent on the specific choices of cut-offs, and similar
effects would have been shown if cut-offs were different.
Discussion
Downloaded from http://hyper.ahajournals.org/ by guest on July 31, 2017
Cardiovascular risk factors are the driving force in the development of cognitive dysfunction,26–28 including Alzheimer’s
disease,29,30 leading to increased prevalence of dementia in the
aging population. Herein, we report that SBP-CV and mean HR
are related independently to cognitive decline as judged from
the MMSE score by a decline ≥5 points, incident cognitive dysfunction defined as clinically relevant cognitive dysfunction to
the MMSE score ≤24, and cognitive deterioration as drop in
MMSE by ≥1 point per year or decline to ≤24 points).
The present analysis investigated the association of SBP and
HR on incident cognitive dysfunction, cognitive decline, and
cognitive deterioration. Many studies were conducted to identify the association of cardiovascular risk with SBP or HR by
using a single measurement of SBP or HR. However, over time,
SBP or HR as risk markers may vary and can place patients
into different risk groups. Time varying statistical analyses
have been used to analyze outcomes by Kaplan–Meier analyses for the shift of patients from one to another risk group,31 an
approach which was used to study the effects of SBP on clinical outcomes in the ONTARGET trial.32 Herein, we have used
multiple readings of HR and SBP for a given patient and averaged values when the number of visits exceeded 2 and determined mean SBP and mean HR over 10.7±2.2 visits. There
was no or only a modest association of mean SBP and mean
HR with the variation parameter, indicating that mean and CV
are independent of each other. The data were collected over
multiple visits and thereby giving a better integration of the
total load of hemodynamic risk markers.
Blood Pressure Mean and Blood Pressure
Variability
SBP was associated with cognitive dysfunction, cognitive decline,
and cognitive deterioration, but after adjusting for confounders,
the effect was not independent of traditional cardiovascular risk
factors. The blood pressure hypothesis has been questioned before
because variation, stability, and episodic signs of hypertension
may be of greater importance in producing cardiovascular outcomes.33 Accordingly, visit to visit variation of SBP was shown
to predict the occurrence of stroke in hypertensive individuals,34
as well as all-cause mortality in the general population35 and was
associated with microbleeds and white matter lesions in patients
after an ischemic stroke.36 Although associations of blood pressure
with cognitive decline are convincing,1,2,37 secondary analyses of
cardiovascular trials aiming to investigate cognitive decline after
blood pressure lowering have been inconsistent. The Syst-Eur
trial included 2418 patients and was terminated prematurely as
a result of significant differences in the incidence of stroke, leading to a short follow-up of only 2 years.38 Dementia incidence
was reduced and there was no significant change of MMSE, but
there was a decline in the control group with decreasing diastolic
blood pressure.38 In patients with prior stroke or transitory ischemic attack, the Perindopril Protection Against Recurrent Stroke
Study (PROGRESS) showed in 6105 randomized patients (with
48% being hypertensive) after follow-up of 3.5 years a relative
risk reduction by 19% of incident dementia or cognitive decline
defined as a MMSE reduction of ≥3 points, but only in patients
with a recurrent stroke.39 In contrast, Systolic Hypertension in
the Elderly Program, the unblinded Medical Research Council’s
Trial on Hypertension in Older Adults, Hypertension in the Very
Elderly Trial, and Study on Cognition and Prognosis in the Elderly
showed trends but no significant effects of hypertensive therapies
on cognitive function.38–43 Also treatment with RAS-inhibitors
and non-RAS inhibitors showed no difference in patients after
stroke (Prevention Regimen for Effectively Avoiding Second
Strokes [PRoFESS])44 or in high risk patients of ONTARGET and
TRANSCEND.24 Interestingly, the association of SBP variation
to stroke7 or cerebral lesions36 has not been investigated in cardiovascular intervention trials in high risk patients, except in one
study on 201 elderly hypertensive individuals in Japan.16 Because
observational studies never proof causality and, in addition, drugs
might act differently on SBP variation,33,34 intervention trials in
high risk patients ought to prospectively investigate pharmacological interventions on SBP variation and outcomes. In a study on
low risk hypertensives, in-trial mean BP was more predictive than
SBP-CV.45 It is open to speculation whether variation of blood
pressure over time is caused by instable blood pressure regulation
adding to the risk of plaque instability.
Heart Rate
In addition to blood pressure, HR has been identified to be associated with cardiovascular outcomes in patients with hypertension,8 myocardial infarction,9 and heart failure.11 Reduction
of clinical events by HR reduction with the If-channel inhibitor ivabradine, qualifying HR as a modifiable risk factor and
not only as a risk marker,10 has only been shown in heart failure.46 In experimental animals after psychosocial stress, pharmacological HR reduction reduces stroke size.14 In patients
after stroke, cognitive decline was accelerated in patients
with a HR >67 bpm and was, furthermore, associated with
mortality, including cardiovascular mortality, but not recurrent stroke rates.15 These findings suggest that stroke quality,
for example, reduction in stroke size rather than a reduction
of second strokes, might occur at low HR.46 Thus, the findings in the PRoFESS trial15 on patients after a previous stroke
can be extended to patients without prior strokes because
there was no significant interaction of HR risk relationship
in patients with or without stroke in this analysis. Therefore,
high HR may be a new risk marker of cognitive dysfunction
and incident dementia in high risk patients with prevalent cardiovascular disease. It is tempting to speculate whether HR
reduction by pharmacological interventions might reduce
risk and delay cognitive dysfunction because in animals with
Böhm et al Blood Pressure, Heart Rate, and Cognitive Decline 659
atherosclerosis, HR reduction has a strong protective effects
on plaque formation,47,48 promotes collateral growth,49 protects
endothelial function, and prevents erectile dysfunction,50 an
association which was also observed in humans.51 Therefore,
HR reduction should be tested in individuals with high risk for
cognitive decline, for example after a stroke. Variation of HR
was not shown to be of predictive value for cognitive decline
or function, and this might be as a result of variability in office
conditions where HR is taken.
Autonomic Regulation in Cognitive Impairment
Downloaded from http://hyper.ahajournals.org/ by guest on July 31, 2017
Changes of HR and blood pressure relation have been observed
in patients with dementia. In patients with Alzheimer’s
dementia, a higher HR was observed compared with controls.52–54 Baroreflex function is impaired in Alzheimer dementia,52 which might contribute to hypotensive syndromes54 and
enhanced postural blood pressure drops with subsequently
impaired cortical perfusion.53 Therefore, autonomic regulation resulting in higher blood pressure variation, greater HR,
and its variability caused by central autonomic dysregulation
might be mechanisms contributing to these findings.
Limitations and Strengths
This analysis is a retrospective exploratory analysis of randomized trials with neutral outcomes. Therefore, the allocation of individuals was not subject to randomization.
However, this is the largest database in patients with high
cardiovascular risk and long follow-up with an average of
>10 visits. Furthermore, this is the first study investigating
the association of variation of SBP over time with mean HR
on incident cognitive decline and incident cognitive dysfunction. The strength is the availability of these measurements
in >10 visits, providing a good integration of HRs over the
whole study period. This analysis is confined to patients in
sinus rhythm because atrial fibrillation per se is associated
with cognitive dysfunction in this cohort.55 Moreover, in
ONTARGET and TRANSCEND, blood pressure was rather
well controlled, and the results could have been different in
patients with uncontrolled or even resistant hypertension.
However, TRANSCEND and ONTARGET are trials on high
risk patients on contemporary treatments for cardiovascular
protection, in whom, on top of proven treatments, SBP-CV
and mean HR show significant associations with cognitive
function and incident impairment of cognition.
Conclusion
This study demonstrates that long-term SBP variations and
mean HR levels are associated with the development of cognitive dysfunction, decline, and deterioration in high risk patients
with atherosclerosis or after stroke or high-risk diabetes mellitus.
Perspectives
The secondary analysis of ONTARGET investigated the SBP
variation and HR on new onset of cognitive decline, dysfunction, and deterioration in a large population of patients at high
cardiovascular risk. HR and blood pressure variation may be
considered markers of cognitive dysfunction in this patient
population. HR and blood pressure variation might be important to judge the future risk of cognitive dysfunction. It should
be incorporated in the clinical judgment of these patients. The
study shows that SBP and mean HR are associated with cognitive dysfunction, with both parameters acting synergistically
on this pathology. They should be incorporated in the work up
of cardiovascular high risk patients. Studies on HR reduction
and smoothening of SBP profiles are warranted.
Sources of Funding
Ongoing Telmisartan Alone and in Combination With Ramipril Global
Endpoint Trial (ONTARGET) and the Telmisartan Randomized
Assessment Study in ACE Intolerant Subjects With Cardiovascular
Disease (TRANSCEND) were funded by Boehringer Ingelheim,
Germany. All authors received scientific support from Boehringer
Ingelheim. M. Böhm and U. Laufs are supported by the Deutsche
Forschungsgesellschaft (KFO 196).
Disclosures
H. Schumacher is an employee of Boehringer Ingelheim, Germany.
The other authors report no conflict.
References
1.Whitmer RA, Sidney S, Selby J, Johnston SC, Yaffe K. Midlife cardiovascular risk factors and risk of dementia in late life. Neurology.
2005;64:277–281. doi: 10.1212/01.WNL.0000149519.47454.F2.
2. Joas E, Bäckman K, Gustafson D, Ostling S, Waern M, Guo X, Skoog I.
Blood pressure trajectories from midlife to late life in relation to dementia
in women followed for 37 years. Hypertension. 2012;59:796–801. doi:
10.1161/HYPERTENSIONAHA.111.182204.
3. Gorelick PB. Role of inflammation in cognitive impairment: results of
observational epidemiological studies and clinical trials. Ann N Y Acad
Sci. 2010;1207:155–162. doi: 10.1111/j.1749-6632.2010.05726.x.
4.Gorelick PB, Scuteri A, Black SE, et al; American Heart Association
Stroke Council, Council on Epidemiology and Prevention, Council on
Cardiovascular Nursing, Council on Cardiovascular Radiology and
Intervention, and Council on Cardiovascular Surgery and Anesthesia.
Vascular contributions to cognitive impairment and dementia: a statement for healthcare professionals from the american heart association/
american stroke association. Stroke. 2011;42:2672–2713. doi: 10.1161/
STR.0b013e3182299496.
5. van Oijen M, de Jong FJ, Witteman JC, Hofman A, Koudstaal PJ, Breteler
MM. Atherosclerosis and risk for dementia. Ann Neurol. 2007;61:403–
410. doi: 10.1002/ana.21073.
6. Pringle E, Phillips C, Thijs L, Davidson C, Staessen JA, de Leeuw PW,
Jaaskivi M, Nachev C, Parati G, O’Brien ET, Tuomilehto J, Webster J,
Bulpitt CJ, Fagard RH; Syst-Eur investigators. Systolic blood pressure
variability as a risk factor for stroke and cardiovascular mortality in the
elderly hypertensive population. J Hypertens. 2003;21:2251–2257. doi:
10.1097/01.hjh.0000098144.70956.0f.
7. Shimbo D, Newman JD, Aragaki AK, LaMonte MJ, Bavry AA, Allison
M, Manson JE, Wassertheil-Smoller S. Association between annual visitto-visit blood pressure variability and stroke in postmenopausal women:
data from the Women’s Health Initiative. Hypertension. 2012;60:625–
630. doi: 10.1161/HYPERTENSIONAHA.112.193094.
8.Kolloch R, Legler UF, Champion A, Cooper-Dehoff RM, Handberg
E, Zhou Q, Pepine CJ. Impact of resting heart rate on outcomes in
hypertensive patients with coronary artery disease: findings from the
INternational VErapamil-SR/trandolapril STudy (INVEST). Eur Heart J.
2008;29:1327–1334. doi: 10.1093/eurheartj/ehn123.
9. Fox K, Ford I, Steg PG, Tendera M, Robertson M, Ferrari R; BEAUTIFUL
investigators. Heart rate as a prognostic risk factor in patients with
coronary artery disease and left-ventricular systolic dysfunction
(BEAUTIFUL): a subgroup analysis of a randomised controlled trial.
Lancet. 2008;372:817–821. doi: 10.1016/S0140-6736(08)61171-X.
10.Reil JC, Custodis F, Swedberg K, Komajda M, Borer JS, Ford I,
Tavazzi L, Laufs U, Böhm M. Heart rate reduction in cardiovascular
disease and therapy. Clin Res Cardiol. 2011;100:11–19. doi: 10.1007/
s00392-010-0207-x.
11. Böhm M, Swedberg K, Komajda M, Borer JS, Ford I, Dubost-Brama
A, Lerebours G, Tavazzi L; SHIFT Investigators. Heart rate as a risk
factor in chronic heart failure (SHIFT): the association between heart
660 Hypertension March 2015
Downloaded from http://hyper.ahajournals.org/ by guest on July 31, 2017
rate and outcomes in a randomised placebo-controlled trial. Lancet.
2010;376:886–894. doi: 10.1016/S0140-6736(10)61259-7.
12. Custodis F, Baumhäkel M, Schlimmer N, List F, Gensch C, Böhm M,
Laufs U. Heart rate reduction by ivabradine reduces oxidative stress,
improves endothelial function, and prevents atherosclerosis in apolipoprotein E-deficient mice. Circulation. 2008;117:2377–2387. doi: 10.1161/
CIRCULATIONAHA.107.746537.
13. Schirmer SH, Degen A, Baumhäkel M, Custodis F, Schuh L, Kohlhaas
M, Friedrich E, Bahlmann F, Kappl R, Maack C, Böhm M, Laufs U.
Heart-rate reduction by If-channel inhibition with ivabradine restores collateral artery growth in hypercholesterolemic atherosclerosis. Eur Heart
J. 2012;33:1223–1231. doi: 10.1093/eurheartj/ehr255.
14. Custodis F, Gertz K, Balkaya M, Prinz V, Mathar I, Stamm C, Kronenberg
G, Kazakov A, Freichel M, Böhm M, Endres M, Laufs U. Heart rate
contributes to the vascular effects of chronic mental stress: effects
on endothelial function and ischemic brain injury in mice. Stroke.
2011;42:1742–1749. doi: 10.1161/STROKEAHA.110.598607.
15. Böhm M, Cotton D, Foster L, Custodis F, Laufs U, Sacco R, Bath PM,
Yusuf S, Diener HC. Impact of resting heart rate on mortality, disability and cognitive decline in patients after ischaemic stroke. Eur Heart J.
2012;33:2804–2812. doi: 10.1093/eurheartj/ehs250.
16. Nagai M, Hoshide S, Ishikawa J, Shimada K, Kario K. Visit-to-visit blood
pressure variations: new independent determinants for cognitive function in the elderly at high risk of cardiovascular disease. J Hypertens.
2012;30:1556–1563. doi: 10.1097/HJH.0b013e3283552735.
17. ONTARGET Investigators, Yusuf S, Teo KK, Pogue J, Dyal L, Copland
I, Schumacher H, Dagenais G, Sleight P, Anderson C. Telmisartan,
ramipril, or both in patients at high risk for vascular events. N Engl J Med.
2008;358:1547–1559.
18. Telmisartan Randomised AssessmeNt Study in ACE iNtolerant subjects
with cardiovascular Disease (TRANSCEND) Investigators, Yusuf S, Teo
K, Anderson C, Pogue J, Dyal L, Copland I, Schumacher H, Dagenais
G, Sleight P. Effects of the angiotensin-receptor blocker telmisartan on
cardiovascular events in high-risk patients intolerant to angiotensinconverting enzyme inhibitors: a randomised controlled trial. Lancet.
2008;372:1174–1183.
19. Hensel A, Luck T, Luppa M, Glaesmer H, Angermeyer MC, Riedel-Heller
SG. Does a reliable decline in Mini Mental State Examination total score
predict dementia? Diagnostic accuracy of two reliable change indices.
Dement Geriatr Cogn Disord. 2009;27:50–58. doi: 10.1159/000189267.
20. Clark CM, Sheppard L, Fillenbaum GG, Galasko D, Morris JC, Koss E,
Mohs R, Heyman A. Variability in annual Mini-Mental State Examination
score in patients with probable Alzheimer disease: a clinical perspective of data from the Consortium to Establish a Registry for Alzheimer’s
Disease. Arch Neurol. 1999;56:857–862.
21. Tombaugh TN, McIntyre NJ. The mini-mental state examination: a comprehensive review. J Am Geriatr Soc. 1992;40:922–935.
22.Hensel A, Angermeyer MC, Riedel-Heller SG. Measuring cognitive
change in older adults: reliable change indices for the Mini-Mental State
Examination. J Neurol Neurosurg Psychiatry. 2007;78:1298–1303. doi:
10.1136/jnnp.2006.109074.
23. Schmand B, Lindeboom J, Launer L, Dinkgreve M, Hooijer C, Jonker C.
What is a significant score change on the Mini-Mental State examination?
Int J Geriatr Psychiatry. 1995;10:411–414.
24.Anderson C, Teo K, Gao P, et al; ONTARGET and TRANSCEND
Investigators. Renin-angiotensin system blockade and cognitive function
in patients at high risk of cardiovascular disease: analysis of data from the
ONTARGET and TRANSCEND studies. Lancet Neurol. 2011;10:43–53.
doi: 10.1016/S1474-4422(10)70250-7.
25. Durrleman S, Simon R. Flexible regression models with cubic splines.
Stat Med. 1989;8:551–561.
26. Solfrizzi V, Panza F, Colacicco AM, D’Introno A, Capurso C, Torres F,
Grigoletto F, Maggi S, Del Parigi A, Reiman EM, Caselli RJ, Scafato
E, Farchi G, Capurso A; Italian Longitudinal Study on Aging Working
Group. Vascular risk factors, incidence of MCI, and rates of progression
to dementia. Neurology. 2004;63:1882–1891.
27.Vermeer SE, Prins ND, den Heijer T, Hofman A, Koudstaal PJ,
Breteler MM. Silent brain infarcts and the risk of dementia and cognitive decline. N Engl J Med. 2003;348:1215–1222. doi: 10.1056/
NEJMoa022066.
28.Kivipelto M, Helkala EL, Laakso MP, Hänninen T, Hallikainen M,
Alhainen K, Soininen H, Tuomilehto J, Nissinen A. Midlife vascular
risk factors and Alzheimer’s disease in later life: longitudinal, population
based study. BMJ. 2001;322:1447–1451.
29.Gorelick PB. Risk factors for vascular dementia and Alzheimer disease. Stroke. 2004;35(11 Suppl 1):2620–2622. doi: 10.1161/01.
STR.0000143318.70292.47.
30.Murray CJL, Lopez AD, eds. The Global Burden of Disease: A
Comprehensive Assessment of Mortality and Disability From Diseases,
Injuries and Risk Factors in 1990 and Projected to 2020. Cambridge, MA:
Havard University Press; 1996.
31. Snappin S, Iglewicz B. Illustrating the impact of a time-varying covariate
with an extended Kaplan-Meier indicator. Am Stat. 2005;4:301–307.
32. Sleight P, Redon J, Verdecchia P, Mancia G, Gao P, Fagard R, Schumacher
H, Weber M, Böhm M, Williams B, Pogue J, Koon T, Yusuf S; ONTARGET
investigators. Prognostic value of blood pressure in patients with high
vascular risk in the Ongoing Telmisartan Alone and in combination with
Ramipril Global Endpoint Trial study. J Hypertens. 2009;27:1360–1369.
doi: 10.1097/HJH.0b013e32832d7370.
33.Rothwell PM. Limitations of the usual blood-pressure hypothesis and
importance of variability, instability, and episodic hypertension. Lancet.
2010;375:938–948. doi: 10.1016/S0140-6736(10)60309-1.
34. Rothwell PM, Howard SC, Dolan E, O’Brien E, Dobson JE, Dahlöf B,
Sever PS, Poulter NR. Prognostic significance of visit-to-visit variability, maximum systolic blood pressure, and episodic hypertension. Lancet.
2010;375:895–905. doi: 10.1016/S0140-6736(10)60308-X.
35. Muntner P, Shimbo D, Tonelli M, Reynolds K, Arnett DK, Oparil S. The
relationship between visit-to-visit variability in systolic blood pressure
and all-cause mortality in the general population: findings from NHANES
III, 1988 to 1994. Hypertension. 2011;57:160–166. doi: 10.1161/
HYPERTENSIONAHA.110.162255.
36. Liu W, Liu R, Sun W, Peng Q, Zhang W, Xu E, Cheng Y, Ding M, Li Y,
Hong Z, Wu J, Zeng J, Yao C, Huang Y; CASISP Study Group. Different
impacts of blood pressure variability on the progression of cerebral
microbleeds and white matter lesions. Stroke. 2012;43:2916–2922. doi:
10.1161/STROKEAHA.112.658369.
37. Birns J, Kalra L. Cognitive function and hypertension. J Hum Hypertens.
2009;23:86–96. doi: 10.1038/jhh.2008.80.
38. Forette F, Seux ML, Staessen JA, et al. Prevention of dementia in randomised double-blind placebo-controlled Systolic Hypertension in
Europe (Syst-Eur) trial. Lancet. 1998;352:1347–1351.
39. Tzourio C, Anderson C, Chapman N, Woodward M, Neal B, MacMahon
S, Chalmers J; PROGRESS Collaborative Group. Effects of blood pressure lowering with perindopril and indapamide therapy on dementia and
cognitive decline in patients with cerebrovascular disease. Arch Intern
Med. 2003;163:1069–1075. doi: 10.1001/archinte.163.9.1069.
40. Applegate WB, Pressel S, Wittes J, Luhr J, Shekelle RB, Camel GH,
Greenlick MR, Hadley E, Moye L, Perry HM Jr. Impact of the treatment of isolated systolic hypertension on behavioral variables. Results
from the systolic hypertension in the elderly program. Arch Intern Med.
1994;154:2154–2160.
41. Prince MJ, Bird AS, Blizard RA, Mann AH. Is the cognitive function of
older patients affected by antihypertensive treatment? Results from 54
months of the Medical Research Council’s trial of hypertension in older
adults. BMJ. 1996;312:801–805.
42.Lithell H, Hansson L, Skoog I, Elmfeldt D, Hofman A, Olofsson B,
Trenkwalder P, Zanchetti A; SCOPE Study Group. The Study on
Cognition and Prognosis in the Elderly (SCOPE): principal results of a
randomized double-blind intervention trial. J Hypertens. 2003;21:875–
886. doi: 10.1097/01.hjh.0000059028.82022.89.
43.Peters R, Beckett N, Forette F, Tuomilehto J, Clarke R, Ritchie C,
Waldman A, Walton I, Poulter R, Ma S, Comsa M, Burch L, Fletcher
A, Bulpitt C; HYVET investigators. Incident dementia and blood
pressure lowering in the Hypertension in the Very Elderly Trial cognitive function assessment (HYVET-COG): a double-blind, placebo controlled trial. Lancet Neurol. 2008;7:683–689. doi: 10.1016/
S1474-4422(08)70143-1.
44. Diener HC, Sacco RL, Yusuf S, et al; Prevention Regimen for Effectively
Avoiding Second Strokes (PRoFESS) study group. Effects of aspirin
plus extended-release dipyridamole versus clopidogrel and telmisartan
on disability and cognitive function after recurrent stroke in patients
with ischaemic stroke in the Prevention Regimen for Effectively
Avoiding Second Strokes (PRoFESS) trial: a double-blind, active and
placebo-controlled study. Lancet Neurol. 2008;7:875–884. doi: 10.1016/
S1474-4422(08)70198-4.
45. Mancia G, Facchetti R, Parati G, Zanchetti A. Visit-to-visit blood pressure variability, carotid atherosclerosis, and cardiovascular events
in the European Lacidipine Study on Atherosclerosis. Circulation.
2012;126:569–578. doi: 10.1161/CIRCULATIONAHA.112.107565.
Böhm et al Blood Pressure, Heart Rate, and Cognitive Decline 661
46. Swedberg K, Komajda M, Böhm M, Borer JS, Ford I, Dubost-Brama A,
Lerebours G, Tavazzi L; SHIFT Investigators. Ivabradine and outcomes
in chronic heart failure (SHIFT): a randomised placebo-controlled study.
Lancet. 2010;376:875–885. doi: 10.1016/S0140-6736(10)61198-1.
47. Custodis F, Schirmer SH, Baumhäkel M, Heusch G, Böhm M, Laufs U.
Vascular pathophysiology in response to increased heart rate. J Am Coll
Cardiol. 2010;56:1973–1983. doi: 10.1016/j.jacc.2010.09.014.
48. Custodis F, Baumhäkel M, Schlimmer N, List F, Gensch C, Böhm M,
Laufs U. Heart rate reduction by ivabradine reduces oxidative stress,
improves endothelial function, and prevents atherosclerosis in apolipoprotein E-deficient mice. Circulation. 2008;117:2377–2387. doi: 10.1161/
CIRCULATIONAHA.107.746537.
49. Schirmer SH, Degen A, Baumhäkel M, Custodis F, Schuh L, Kohlhaas
M, Friedrich E, Bahlmann F, Kappl R, Maack C, Böhm M, Laufs U.
Heart-rate reduction by If-channel inhibition with ivabradine restores collateral artery growth in hypercholesterolemic atherosclerosis. Eur Heart
J. 2012;33:1223–1231. doi: 10.1093/eurheartj/ehr255.
50. Baumhäkel M, Custodis F, Schlimmer N, Laufs U, Böhm M. Heart rate reduction with ivabradine improves erectile dysfunction in parallel to decrease
in atherosclerotic plaque load in ApoE-knockout mice. Atherosclerosis.
2010;212:55–62. doi: 10.1016/j.atherosclerosis.2010.03.002.
51.Kratz MT, Schumacher H, Sliwa K, Schmieder R, Pöss J, Mahfoud
F, Unger T, Lonn E, Koon T, Mancia G, Sleight P, Yusuf S, Böhm M.
Heart rate and blood pressure interactions in the development of erectile dysfunction in high-risk cardiovascular patients. Eur J Prev Cardiol.
2014;21:272–280. doi: 10.1177/2047487313494835.
52. Meel-van den Abeelen AS, Lagro J, Gommer ED, Reulen JP, Claassen
JA. Baroreflex function is reduced in Alzheimer’s disease: a candidate biomarker? Neurobiol Aging. 2013;34:1170–1176. doi: 10.1016/j.
neurobiolaging.2012.10.010.
53.van Beek AH, Sijbesma JC, Jansen RW, Rikkert MG, Claassen JA.
Cortical oxygen supply during postural hypotension is further decreased
in Alzheimer’s disease, but unrelated to cholinesterase-inhibitor use. J
Alzheimers Dis. 2010;21:519–526. doi: 10.3233/JAD-2010-100288.
54. Schoon Y, Lagro J, Verhoeven Y, Rikkert MO, Claassen J. Hypotensive
syndromes are not associated with cognitive impairment in geriatric patients. Am J Alzheimers Dis Other Demen. 2013;28:47–53. doi:
10.1177/1533317512466692.
55. Marzona I, O’Donnell M, Teo K, Gao P, Anderson C, Bosch J, Yusuf S.
Increased risk of cognitive and functional decline in patients with atrial
fibrillation: results of the ONTARGET and TRANSCEND studies. CMAJ.
2012;184:E329–E336. doi: 10.1503/cmaj.111173.
Downloaded from http://hyper.ahajournals.org/ by guest on July 31, 2017
Novelty and Significance
What Is New?
Summary
• The secondary analysis of Ongoing Telmisartan Alone and in Combina-
The study shows that systolic blood pressure and mean heart rate
are associated with cognitive dysfunction, with both parameters
acting synergistically on this pathology. They should be incorporated in the work up of cardiovascular high risk patients. Studies
on heart rate reduction and smoothening of systolic blood pressure
profiles are warranted.
tion With Ramipril Global Endpoint Trial investigated the systolic blood
pressure variation and heart rate on new onset of cognitive decline,
dysfunction, and deterioration in a large population of patients at high
cardiovascular risk.
• Heart rate and blood pressure variation may be considered as markers of
cognitive dysfunction in this patient population.
What Is Relevant?
• Heart rate and blood pressure variation might be important to judge the
future risk of cognitive dysfunction. It should be incorporated in the clinical judgment of these patients.
Downloaded from http://hyper.ahajournals.org/ by guest on July 31, 2017
Systolic Blood Pressure Variation and Mean Heart Rate Is Associated With Cognitive
Dysfunction in Patients With High Cardiovascular Risk
Michael Böhm, Helmut Schumacher, Darryl Leong, Giuseppe Mancia, Thomas Unger, Roland
Schmieder, Florian Custodis, Hans-Christoph Diener, Ulrich Laufs, Eva Lonn, Karen Sliwa,
Koon Teo, Robert Fagard, Josep Redon, Peter Sleight, Craig Anderson, Martin O'Donnell and
Salim Yusuf
Hypertension. 2015;65:651-661; originally published online January 12, 2015;
doi: 10.1161/HYPERTENSIONAHA.114.04568
Hypertension is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231
Copyright © 2015 American Heart Association, Inc. All rights reserved.
Print ISSN: 0194-911X. Online ISSN: 1524-4563
The online version of this article, along with updated information and services, is located on the
World Wide Web at:
http://hyper.ahajournals.org/content/65/3/651
Data Supplement (unedited) at:
http://hyper.ahajournals.org/content/suppl/2015/01/12/HYPERTENSIONAHA.114.04568.DC1
http://hyper.ahajournals.org/content/suppl/2016/04/11/HYPERTENSIONAHA.114.04568.DC2
Permissions: Requests for permissions to reproduce figures, tables, or portions of articles originally published
in Hypertension can be obtained via RightsLink, a service of the Copyright Clearance Center, not the Editorial
Office. Once the online version of the published article for which permission is being requested is located,
click Request Permissions in the middle column of the Web page under Services. Further information about
this process is available in the Permissions and Rights Question and Answer document.
Reprints: Information about reprints can be found online at:
http://www.lww.com/reprints
Subscriptions: Information about subscribing to Hypertension is online at:
http://hyper.ahajournals.org//subscriptions/
HYPE10140568
28.11.2014
Systolic blood pressure variation and mean heart rate is associated
with cognitive dysfunction in patients with high cardiovascular risk
ONLINE SUPPLEMENT
Michael Böhm1, Helmut Schumacher2, Darryl Leong3, Giuseppe Mancia4, Thomas
Unger5, Roland Schmieder6, Florian Custodis1, Hans-Christoph Diener7, Ulrich
Laufs1, Eva Lonn3, Karen Sliwa8, Koon Teo3, Robert Fagard9, Josep Redon10, Peter
Sleight11, Craig Anderson12, Martin O’Donnell13, Salim Yusuf3
Correspondence to:
Michael Böhm, MD
Universitätsklinikum des Saarlandes
Klinik für Innere Medizin III
Kardiologie, Angiologie und Internistische Intensivmedizin
Kirrberger Str. 1
D 66424 Homburg/Saar
Germany
Tel.: (+49)-6841-16-23372
Fax: (+49)-6841-16-23369
E-mail: [email protected]
1
2
3
4
5
6
7
8
9
10
11
12
13
Böhm M, Custodis F, Laufs U: Klinik für Innere Medizin III, Universitätsklinikum des
Saarlandes, DE-66424 Homburg, Germany
Schumacher H: Boehringer Ingelheim, Pharma GmbH & Co. KG, DE-55216 Ingelheim,
Germany
Leong D, Lonn E, Teo K, Yusuf S: Population Health Research Institute, McMaster
University, Hamilton, Ontario L8L 2X2, Canada
Mancia G: Centro di Fisiologica Clinica e Ipertensione, Universita Milano-Bicocca, Istituto
Auxologico, Milan, Italy
Unger T: CARIM-School for Cardiovascular Diseases, Maastricht University, Maastricht,
The Netherlands
Schmieder R: Department of Nephrology and Hypertension, Friedrich-Alexander
University, DE-91054 Erlangen, Germany
Diener HC: Department of Neurology, University Hospital Essen, DE-45122 Essen,
Germany
Sliwa K: Hatter Institute for Cardiovascular Research in Africa & IIDMM; Faculty of Health
Sciences; University of Cape Town, South Africa
Fagard R: Hypertension Unit, KU Leuven University, B-3000 Leuven, Belgium
Redon J: University of Valencia, Spain
Sleight P: Department of Cardiovascular Medicine, John Radcliffe Hospital, Oxford, OX3
9DU, United Kingdom
Anderson C: The George Institute for Global Health, University of Sydney and Royal
Prince Alfred Hospital, Sydney, NSW, Australia
O’Donnell M: HRB Clinical Research Facility Galway, National University of Ireland,
Galway, Geata an Eolais, University Road, Galway, Ireland.
Running Title: Blood pressure, heart rate and cognitive decline
1
table S1: Demographics and Baseline Conditions
Included in Analysis
No
A: SBP mean quintiles
1
≤123.7 mmHg
Variable
Unit
Yes
Patients
N
Available visits
Mean (Range)
Mean SBP [mmHg]
Mean (SD)
SBP variability (CV) [%]
Mean (SD)
10.1 (3.7)
9.6 (3.3)
Mean heart rate [bpm]
Mean (SD)
70.4 (8.6)
Heart rate variability (CV) [%]
Mean (SD)
Mean DBP [mmHg]
2
>123.7-≤131.2
mmHg
3
131.2-≤137.6
mmHg
B: SBP CV quintiles
4
>137.6-≤145.3
mmHg
5
>145.3 mmHg
1
≤6.8%
2
6.8-8.4%
3
8.4-10.0%
C: HR mean quintiles
4
10.0-12.2%
5
≥12.2%
1
≤61.3 bpm
2
61.3-66.3 bpm
3
66.3-70.6 bpm
D: HR CV quintiles
4
70.7-75.8 bpm
5
≥75.8 bpm
1
≤7.1 %
2
>7.1-≤9.0 %
3
>9.0-≤10.9 %
4
>10.9-≤13.8 %
2
>13.8 %
6953
24593
4920
4922
4904
4923
4924
4918
4919
4919
4919
4918
4924
4904
4909
4931
4925
4918
4919
4919
4919
4918
5.3 (0-13)
10.7 (3-13)
10.8 (3-13)
10.8 (3-13)
10.8 (3-13)
10.7 (3-13)
10.5 (3-13)
10.5 (3-13)
10.9 (3-13)
10.9 (3-13)
10.9 (3-13)
10.5 (3-13)
10.8 (3-13)
10.8 (3-13)
10.8 (3-13)
10.7 (3-13)
10.4 (3-13)
10.3 (3-13)
10.9 (3-13)
10.9 (3-13)
10.9 (3-13)
10.6 (3-13)
137.6 (13.7) 134.7 (13.2)
116.6 (5.7)
127.7 (2.1)
134.4 (1.8)
141.2 (2.2)
153.6 (7.4)
134.9 (12.1)
134.8 (12.9)
134.1 (13.3)
134.6 (13.7)
135.1 (14.1)
135.0 (14.0)
134.4 (13.4)
134.0 (12.6)
134.7 (12.7)
135.4 (13.3)
134.8 (13.2)
134.2 (12.8)
134.2 (13.0)
134.8 (13.3)
135.5 (13.8)
9.8 (3.3)
9.4 (3.3)
9.3 (3.3)
9.4 (3.3)
9.8 (3.3)
5.4 (1.1)
7.6 (0.5)
9.2 (0.5)
11 (0.6)
14.6 (2.3)
9.9 (33)
9.6 (3.3)
9.4 (3.3)
9.3 (3.3)
9.6 (3.4)
8.9 (3.4)
9.2 (3.2)
9.5 (3.2)
9.8 (3.3)
10.4 (3.4)
68.7 (8.8)
68.2 (8.6)
68.7 (8.5)
69.0 (8.7)
68.9 (8.7)
68.6 (9.3)
69.2 (8.4)
68.9 (8.5)
68.6 (8.7)
68.4 (9.0)
68.5 (9.2)
56.8 (3.5)
63.8 (1.4)
68.5 (1.3)
73.1 (1.5)
81.2 (4.9)
68.2 (8.6)
68.4 (8.8)
69.0 (8.6)
69.2 (8.8)
68.7 (9.0)
11.1 (5.3)
10.7 (4.8)
10.6 (4.5)
10.6 (4.9)
10.6 (4.7)
10.7 (5.1)
11.1 (5.1)
9.6 (4.6)
10.4 (4.3)
10.8 (4.7)
11.2 (5.0)
11.7 (5.3)
10.6 (4.5)
10.8 (4.5)
10.6 (4.6)
10.5 (4.7)
11.2 (5.8)
5.7 (1.2)
8.1 (0.6)
9.9 (0.6)
12.2 (0.8)
17.8 (5.4)
Mean (SD)
77.9 (7.8)
77.1 (7.5)
70.1 (5.6)
75.0 (5.4)
77.6 (5.7)
79.8 (6.0)
83.2 (7.1)
78.3 (6.8)
77.5 (7.2)
77 (7.5)
76.7 (7.6)
76.2 (8.0)
75.2 (7.6)
76.6 (7.5)
77.0 (7.2)
77.9 (7.1)
79.0 (7.5)
77.1 (7.3)
77.0 (7.3)
77.0 (7.4)
77.3 (7.6)
77.3 (7.7)
DBP variability (CV) [%]
Mean (SD)
10.6 (3.8)
10.2 (3.5)
10.8 (3.5)
10.1 (3.6)
9.9 (3.4)
9.9 (3.5)
10.1 (3.5)
7.6 (2.7)
9.0 (2.7)
10.0 (2.9)
11.1 (2.9)
13.1 (3.5)
10.7 (3.5)
10.2 (3.5)
10.0 (3.4)
9.8 (3.5)
10.2 (3.6)
9.3 (3.4)
9.8 (3.4)
10.1 (3.4)
10.5 (3.4)
11.1 (3.6)
MMSE
Mean (SD)
24.2 (4.3)
28.5 (1.5)
28.6 (1.5)
28.6 (1.5)
28.5 (1.5)
28.5 (1.5)
28.3 (1.6)
28.6 (1.5)
28.6 (1.5)
28.5 (1.5)
28.5 (1.5)
28.4 (1.6)
28.5 (1.5)
28.5 (1.5)
28.6 (1.5)
28.5 (1.5)
28.4 (1.6)
28.6 (1.5)
28.5 (1.5)
28.6 (1.5)
28.5 (1.5)
28.4 (1.5)
Age [years]
Mean (SD)
68.3 (7.6)
66.0 (7.0)
64.3 (6.8)
65.5 (7.1)
65.9 (7.0)
66.6 (7.0)
67.6 (7.0)
65.1 (6.9)
65.4 (6.9)
65.9 (6.9)
66.4 (7.0)
67.2 (7.3)
66.9 (6.8)
66.3 (6.9)
66 (7.1)
65.7 (7.1)
65.2 (7.1)
66.0 (7.1)
65.6 (7.0)
65.7 (7.0)
66.1 (7.1)
66.6 (7.1)
Body mass index [kg/m2]
Mean (SD)
28.1 (5.0)
28.2 (4.7)
27.5(4.5)
28.1 (4.6)
28.3 (4.9)
28.5 (4.7)
28.5 (4.7)
28.4 (4.7)
28.2 (4.5)
28.3 (4.7)
28.1 (4.8)
28.0 (4.9)
27.8 (4.2)
28.0 (4.5)
28.1 (4.6)
28.4 (5.0)
28.6 (5.1)
28.2 (4.6)
28.1 (4.6)
28.2 (4.7)
28.2 (4.8)
28.3 (4.9)
eGFR (MDRD) [ml/min/1.73 m2]
Mean (SD)
71.4 (22.0)
74.0 (19.4)
76.2 (18.2)
75.1 (19.5)
74.2 (19.7)
73.2 (19.7)
71.5 (19.6)
74.5 (18.8)
75.2 (19.0)
74.4 (19.3)
73.6 (19.7)
72.5 (20.3)
73.0 (17.5)
73.6 (18.5)
74.3 (20.1)
74.1 (20.4)
75.1 (20.4)
74.9 (19.7)
74.2 (19.4)
74.5 (19.6)
73.8 (19.3)
72.7 (19.0)
Male
%
62.4
72.5
78.2
76.6
71.7
70.1
65.8
74.1
75.3
73.3
71.3
68.5
81.7
74.7
71.6
67.4
67.1
70.2
70.4
72.1
74.0
75.8
Asian
%
13.7
15.6
17.4
16.3
16.5
15.3
12.4
13.7
14.7
16.0
17.1
16.4
10.2
13.5
15.5
18.5
20.1
13.9
15.9
16.0
16.2
15.7
Black
%
4.1
1.8
1.3
1.4
1.5
2.2
2.7
0.9
1.4
1.9
1.7
3.3
1.5
1.6
1.8
1.6
2.6
1.6
1.6
1.5
1.6
2.8
Other
%
15.9
8.9
7.9
9.2
9.2
9.3
9.2
8.2
8.4
9.1
9.5
9.5
7.1
8.6
8.6
9.9
10.4
7.3
8.6
8.7
9.6
10.5
White
%
66.2
73.7
73.4
73.1
72.8
73.3
75.7
77.2
75.6
73.0
71.7
70.9
81.2
76.2
74.1
70.0
66.9
77.1
73.9
73.8
72.6
71.0
Mainly sedentary
%
33.8
20.1
16.9
19.2
20.8
21.3
22.4
21.0
18.5
19.2
19.8
22.1
14.3
17.9
20.8
22.1
25.5
19.6
19.6
20.1
20.5
20.8
< Once/week
%
13.4
11.0
10.1
11.0
11.5
10.8
11.5
11.5
10.7
11.2
10.5
11.0
8.9
11.1
10.3
12.4
12.2
10.6
10.1
11.5
11.2
11.6
2-4 times/week
%
19.2
23.8
24.3
24.8
24.0
23.3
22.7
23.8
24.3
23.3
24.2
23.5
25.0
25.9
24.3
22.5
21.5
23.8
24.2
23.9
23.4
23.9
5-6 times/week
%
6.1
8.0
8.9
8.0
8.0
7.8
7.4
8.7
7.7
8.7
7.9
7.2
10.1
8.3
8.2
7.3
6.2
8.5
8.2
8.2
7.6
7.6
Everyday
%
27.1
37.0
39.6
37.0
35.8
36.8
36.0
35.1
38.8
37.6
37.6
36.1
41.6
36.7
36.4
35.8
34.6
37.5
37.9
36.3
37.3
36.1
None
%
10.3
1.8
1.2
1.6
1.8
2.2
2.3
1.8
1.9
1.7
1.7
1.9
1.3
1.5
1.7
2.1
2.4
1.9
1.8
1.5
2.1
1.8
1-8 years
%
40.8
27.1
23.5
24.7
26.8
28.5
31.9
25.4
25.0
27.1
27.5
30.3
24.2
25.9
26.8
27.7
30.6
25.5
26.7
26.7
27.7
28.7
9-12 years
%
25.7
30.6
31.9
31.6
30.6
29.7
29.2
28.8
29.3
30.7
31.6
32.6
31.6
30.1
30.7
29.4
31.2
30.4
30.4
30.7
30.4
31.1
Trade/Tech school
%
12.2
19.4
19.7
19.5
19.1
19.1
19.7
20.4
21.1
19.4
18.9
17.3
21.6
20.1
18.5
19.6
17.2
19.9
19.9
19.8
18.7
18.7
College/Univ.
%
10.7
21.1
23.8
22.7
21.7
20.6
16.9
23.6
22.7
21.2
20.3
17.9
21.3
22.5
22.3
21.2
18.4
22.3
21.2
21.4
21.1
19.7
%
29.2
41.4
43.5
42.7
40.3
41.2
39.2
40.9
43.2
41.9
42.3
38.6
49.7
43.6
41.1
36.7
35.7
42.8
41.4
40.7
40.8
41.1
Current
%
12.3
12.0
14.6
12.6
11.5
11.5
9.9
10.9
11.8
12.1
11.6
13.6
8.6
10.1
11.6
12.9
16.9
12.2
12.2
12.9
12.0
10.8
Formerly
%
44.6
51.8
54.4
53.9
51.2
50.4
48.8
48.9
52.3
52.0
52.8
52.8
61.0
55.1
50.4
47.0
45.4
50.7
49.8
50.3
53.1
55.0
Never
%
42.4
36.2
31.0
33.4
37.4
38.1
41.2
40.2
35.9
35.9
35.6
33.6
30.4
34.8
38.1
40.1
37.7
37.1
38.0
36.9
34.9
34.3
Depression
%
23.8
21.0
23.2
20.1
20.1
19.8
21.5
18.3
19.9
21.2
22.0
23.4
21.0
20.3
19.9
20.1
23.5
20.5
21.2
20.1
20.8
22.2
History of hypertension
%
74.2
69.0
43.3
63.2
72.5
79.9
86.2
68.6
68.0
67.7
69.1
71.6
63.3
67.0
69.6
71.7
73.4
67.7
69.0
68.6
69.2
70.6
History of diabetes
%
42.3
35.7
24.2
31.8
36.5
40.2
46.0
36.4
35.3
35.0
34.3
37.7
22.7
28.4
33.9
41.6
52.0
37.8
37.0
35.0
34.3
34.6
New diabetes
%
5.2
10.7
10.1
12.0
10.7
10.4
10.3
10.0
10.7
10.9
11.3
10.6
11.1
11.8
11.1
10.4
8.9
9.4
10.2
11.1
10.8
12.0
Previous Myocardial infarction
%
45.6
49.3
60.6
53.2
48.7
44.7
39.2
49.4
50.1
49.8
49.7
47.5
58.3
54.0
51.6
45.7
36.9
48.6
48.4
49.8
50.5
49.1
Previous Stroke/TIA
%
26.2
19.6
15.5
18.2
19.9
22.0
22.3
18.1
19.3
19.2
19.8
21.7
14.3
17.0
19.9
22.4
24.4
18.6
19.4
19.0
20.0
20.9
New stroke
%
1.8
3.0
1.7
2.8
2.7
3.6
4.0
2.5
2.5
2.6
3.3
3.9
2.5
2.6
2.5
3.1
4.0
2.4
2.6
2.9
3.2
3.7
Atrial fibrillation
%
4.8
2.9
3.1
3.1
2.6
2.7
3.0
2.8
2.5
2.6
3.2
3.4
1.5
1.6
2.3
3.6
5.5
1.6
1.5
2.2
3.4
5.8
New AF
%
3.0
6.2
5.2
5.6
6.5
6.7
6.9
4.6
5.4
6.0
6.9
8.0
6.2
6.4
6.9
5.8
5.5
3.1
3.6
4.7
6.0
13.5
ASA
%
72.0
76.5
81.8
78.3
75.9
74.0
72.7
74.9
78.0
78.0
76.4
75.5
84.1
81.4
77.9
71.7
67.6
75.4
77.8
77.0
76.4
76.1
Beta-blockers
%
52.6
58.5
61.7
58.1
58.4
56.7
57.4
56.8
56.8
58.1
59.4
61.2
80.9
68.5
59.2
48.7
35.1
57.0
56.9
57.1
58.8
62.5
Diuretics
%
36.6
26.7
19.2
23.5
26.0
29.8
35.1
24.4
24.5
26.7
26.5
31.6
23.0
24.6
26.8
29.2
30.1
25.6
24.8
26.2
26.7
30.4
Nitrates
%
31.2
30.0
33.5
32.5
31.7
28.5
23.8
27.6
28.3
29.3
31.4
33.1
32.3
32.6
31.9
28.6
24.6
29.0
29.5
30.0
29.9
31.6
CC blockers
%
25.4
25.1
15.8
22.1
26.5
29.0
32.1
26.5
25.3
25.1
24.5
24.1
23.4
25.2
25.4
25.7
25.8
25.7
24.6
24.8
25.0
25.3
Oral hypoglycemic agents
%
27.9
24.0
16.4
20.9
25.0
27.2
30.6
25.4
23.1
23.6
23.2
24.8
14.5
18.1
22.6
28.5
36.4
25.6
25.1
23.3
23.0
23.1
Statins
%
52.9
62.5
71.3
64.8
62.1
58.4
56.1
58.1
62.5
63.9
63.8
64.4
75.5
67.9
61.8
56.5
51.0
62.4
61.9
61.8
62.7
63.9
Race
Physical activity
Formal education
Alcohol consumption
Tobacco use
Comorbidities
Regular medication
table S2 a Additive effects of SBP CV and HR mean on cognitive dysfunction (MMSE>=24), odds ratios with 95% CIs yellow: odds increased by at least 10%; orange: odds increased by at least 25%; red: odds increased by at least 50% (all based on lower confidence boundary) SBP CV quintile 1st HR mean quintile 1st 2nd 3rd 4th 1 1.23 1.35 1.38 (reference) (1.04-­‐1.46) (1.14-­‐1.60) (1.16-­‐1.63) 2nd 1.09 1.34 1.46 1.50 (0.91-­‐1.29) (1.05-­‐1.71) (1.15-­‐1.86) (1.17-­‐1.91) 3rd 0.98 1.20 1.32 1.35 (0.82-­‐1.17) (0.94-­‐1.55) (1.03-­‐1.69) (1.05-­‐1.73) 4th 1.21 1.49 1.63 1.67 (1.01-­‐1.45) (1.16-­‐1.91) (1.27-­‐2.09) (1.30-­‐2.14) 5th 1.32 1.62 1.78 1.82 (1.10-­‐1.58) (1.26-­‐2.09) (1.38-­‐2.29) (1.41-­‐2.34) Test on SBP-­‐CV by HR mean interaction: p=0.55 table S2 b Additive effects of SBP CV and HR mean on cognitive decline (reduction in MMSE>=5), odds ratios with 95% CIs SBP CV quintile 1st HR mean quintile 1st 2nd 3rd 4th 1 1.18 1.53 1.53 (reference) (0.95-­‐1.46) (1.25-­‐1.88) (1.25-­‐1.89) 2nd 1.07 1.27 1.65 1.65 (0.88-­‐1.32) (0.94-­‐1.70) (1.23-­‐2.20) (1.23-­‐2.21) 3rd 1.02 1.21 1.57 1.57 (0.83-­‐1.26) (0.90-­‐1.63) (1.17-­‐2.10) (1.17-­‐2.11) 4th 1.07 1.26 1.63 1.63 (0.86-­‐1.31) (0.93-­‐1.70) (1.21-­‐2.20) (1.21-­‐2.20) 5th 1.33 1.56 2.03 2.03 (1.07-­‐1.64) (1.15-­‐2.12) (1.50-­‐2.74) (1.50-­‐2.75) Test on SBP-­‐CV by HR mean interaction: p=0.11 table S2 c 5th 1.40 (1.18-­‐1.66) 1.52 (1.19-­‐1.94) 1.37 (1.07-­‐1.76) 1.70 (1.33-­‐2.18) 1.85 (1.44-­‐2.38) 5th 1.44 (1.16-­‐1.78) 1.54 (1.15-­‐2.08) 1.47 (1.09-­‐1.98) 1.53 (1.13-­‐2.07) 1.91 (1.40-­‐2.58) Additive effects of SBP CV and HR mean on cognitive impairment/decline, odds ratios with 95% CIs SBP CV quintile 1st HR mean quintile 1st 2nd 3rd 4th 1 1.17 1.36 1.35 (reference) (1.00-­‐1.37) (1.17-­‐1.59) (1.15-­‐1.57) 2nd 1.10 1.28 1.50 1.48 (0.94-­‐1.28) (1.03-­‐1.60) (1.21-­‐1.86) (1.19-­‐1.85) 3rd 1.00 1.17 1.38 1.35 (0.86-­‐1.18) (0.94-­‐1.47) (1.09-­‐1.71) (1.08-­‐1.70) 4th 1.20 1.40 1.63 1.62 (1.02-­‐1.41) (1.12-­‐1.76) (1.31-­‐2.04) (1.29-­‐2.03) 5th 1.37 1.60 1.87 1.85 (1.17-­‐1.62) (1.27-­‐2.02) (1.49-­‐2.35) (1.47-­‐2.33) Test on SBP-­‐CV by HR mean interaction: p=0.70 5th 1.44 (1.23-­‐1.68) 1.59 (1.27-­‐1.97) 1.45 (1.16-­‐1.81) 1.73 (1.38-­‐2.16) 1.98 (1.58-­‐2.49) table S3
Sensitivity analysis in a linear model for predictors for reduction of MMSE per year;
variables from logistic regression model selection (cf. Table 1) with p<0.15 presented
(global p-value given for predictors with more than 2 categories) Variables
Estimate
SBP CV quintiles
95% CI
p<.0001
2 vs 1
0.007
(-0.017 - 0.032)
3 vs 1
0.006
(-0.019 - 0.031)
4 vs 1
0.031
(0.005 - 0.057)
5 vs 1
0.062
(0.034 - 0.090)
Mean HR quintiles
p=0.0018
2 vs 1
-0.008
(-0.032 - 0.016)
3 vs 1
0.025
(0.000 - 0.049)
4 vs 1
0.031
(0.006 - 0.055)
5 vs 1
0.033
(0.008 - 0.059)
0.123
(0.118 - 0.129)
MMSE at baseline, per score unit
DBP CV quintiles
p=0.022
2 vs 1
-0.030
(-0.054 - -0.005)
3 vs 1
-0.040
(-0.065 - -0.014)
4 vs 1
-0.020
(-0.047 - 0.006)
5 vs 1
-0.014
(-0.042 - 0.014)
Age, per year
0.012
(0.011 - 0.013)
Female vs Male
0.015
(-0.003 - 0.034)
Ethnicity
p<.0001
Asian vs White
-0.020
(-0.043 - 0.003)
Black vs White
0.154
(0.096 - 0.211)
Other vs White
0.045
(0.017 - 0.074)
Physical activity
p<.0001
Everyday vs Mainly sedentary
-0.064
(-0.086 - -0.043)
5-6 times/week vs Mainly sedentary
-0.050
(-0.082 - -0.018)
2-4 times/week vs Mainly sedentary
-0.046
(-0.069 - -0.022)
≤ Once/week vs Mainly sedentary
-0.036
(-0.065 - -0.008)
Formal education
p<0.0001
College/University vs ≤ 8 years
-0.168
(-0.191 - -0.146)
Trade/Technical vs ≤ 8 years
-0.124
(-0.147 - -0.101)
9-12 years vs ≤ 8 years
-0.100
(-0.120 - -0.080)
Alcohol consumption
-0.031
(-0.048 - -0.014)
History of stroke
0.045
(0.024 - 0.065)
New stroke
0.294
(0.249 - 0.340)
History of diabetes
0.050
(0.024 - 0.074)
Nitrates
0.018
(0.001 - 0.035)
Statins
-0.016
(-0.033 - 0.001)
Hypoglycemics
-0.022
(-0.050 - 0.005)
Concomitant medication
figure S1 Cumulative percentage of patients below cut-­‐off levels for (A) final MMSE by SBP-­‐CV quintiles, (B) yearly MMSE change by SBP-­‐CV quintiles, (C) final MMSE by HR mean quintiles, (D) yearly MMSE change by HR mean quintiles A B C D figure S1
Cumulative percentage of patients below MMSE at time of last follow-up (A, C) and below
change in MMSE per year (B, D) by SBP-CV quintiles (A, B) and mean HR quintiles (C, D).
It is shown that the highest quintiles consistently show highest percentages across all
values on the abscissa (x-axis).This indicates that the detrimental effect of high SBP-CV
and mean HR is independent of the choice of cut-off levels (MMSE at last follow-up of 25,
yearly reduction of 1 point).