Accepted Manuscript

Hemoglobin level and three-month clinical outcomes among

ischemic stroke patients with elevated systolic blood pressure

Daoxia Guo, Zhengbao Zhu, Chongke Zhong, Hao Peng, Tian Xu,
Aili Wang, Yanbo Peng, Tan Xu, Chung-Shiuan Chen, Yongqiu
Li, Zhong Ju, Jing Chen, Yonghong Zhang, Jiang He

PII: S0022-510X(18)30483-0
DOI: https://doi.org/10.1016/j.jns.2018.11.030
Reference: JNS 16104
To appear in: Journal of the Neurological Sciences
Received date: 8 October 2018
Revised date: 19 November 2018
Accepted date: 26 November 2018

Please cite this article as: Daoxia Guo, Zhengbao Zhu, Chongke Zhong, Hao Peng, Tian
Xu, Aili Wang, Yanbo Peng, Tan Xu, Chung-Shiuan Chen, Yongqiu Li, Zhong Ju, Jing
Chen, Yonghong Zhang, Jiang He , Hemoglobin level and three-month clinical outcomes
among ischemic stroke patients with elevated systolic blood pressure. Jns (2018),
https://doi.org/10.1016/j.jns.2018.11.030

This is a PDF file of an unedited manuscript that has been accepted for publication. As
a service to our customers we are providing this early version of the manuscript. The
manuscript will undergo copyediting, typesetting, and review of the resulting proof before
it is published in its final form. Please note that during the production process errors may
be discovered which could affect the content, and all legal disclaimers that apply to the
journal pertain.
 ACCEPTED MANUSCRIPT

Hemoglobin level and three-month clinical outcomes among ischemic stroke

patients with elevated systolic blood pressure

Daoxia Guo MDa；Zhengbao Zhu, MDa, b；Chongke Zhong, MDa, b；Hao Peng, MD,

PhDa；Tian Xu, MD, PhDa, c；Aili Wang, MD, PhDa；Yanbo Peng, MD PhDd；Tan Xu,

MD, PhDa；Chung-Shiuan Chen, MSb ；Yongqiu Li, MD, PhDe ；Zhong Ju, MD, PhDf；

PT
Jing Chen, MD, MSb, g；Yonghong Zhang, MD, PhDa# ；Jiang He, MD, PhDb,g

RI
SC
a
Department of Epidemiology, School of Public Health and Jiangsu Key Laboratory

of Preventive and Translational Medicine for Geriatric Diseases, Medical College of

NU
Soochow University, Suzhou, Jiangsu, China.
MA

b
Department of Epidemiology, Tulane University School of Public Health and

Tropical Medicine, New Orleans, LA

ED

c
Department of Neurology, Affiliated Hospital of Nantong University, Nantong,
T

Jiangsu, China
EP

d
Department of Neurology, Affiliated Hospital of Hebei United University, Hebei,

China
C
AC

e
Department of Neurology, Tangshan Worker’s Hospital, Hebei, China

f
Department of Neurology, Kerqin District First People’s Hospital of Tongliao City,

Inner Mongolia, China

g
Department of Medicine, Tulane University School of Medicine, New Orleans, LA,

USA

1
 ACCEPTED MANUSCRIPT

#
Correspondence to:

Yonghong Zhang, MD, PhD,

School of Public Health, Medical College of Soochow University, 199 Renai Road,

Industrial Park District, Suzhou, 215123, China

Telephone: +86-512-6588-0078

PT
Fax: +86-512-6588-0052

RI
E-mail: [email protected]

SC
NU
MA
T ED
C EP
AC

2
 ACCEPTED MANUSCRIPT

Abstract

Background: Previous studies have reported that extreme low and high hemoglobin

levels are positively associated with the risk of ischemic stroke. However, there are

few reports on the relationship between hemoglobin at acute phase and clinical

outcomes after ischemic stroke and the results of their association to date are

PT
inconsistent. We aimed to investigate the association between them in a large

RI
prospective cohort of ischemic stroke patients.

SC
Methods: Baseline hemoglobin levels were measured in 3881 patients with acute

ischemic stroke. The primary outcome was defined as composite outcome of major
NU
disability and death (modified Rankin Scale score ≥3) at 3 months after stroke onset.
MA

Secondary outcomes were separately those of major disability and death.

Results: Compared with the lowest quartile of hemoglobin, the multivariate adjusted
ED

odds ratios (95% confidence intervals) associated with the highest quartile were 1.38
T

(1.03-1.86), 1.49 (1.11-1.99), 0.79 (0.41-1.52) for primary outcome, major disability
EP

and death, respectively. Multiple-adjusted spline regression model showed linear

associations of hemoglobin levels with primary outcome (P for linearity =0.037) and
C
AC

major disability (P for linearity =0.004). Subgroup analyses further confirmed the

positive association between high hemoglobin and poor prognosis of ischemic stroke.

Conclusions: Elevated hemoglobin levels in the acute phase were associated with

poor prognosis at 3 months after ischemic stroke. Further prospective studies from

other samples of ischemic stroke patients are needed to validate our findings.

Keywords: hemoglobin; ischemic stroke; prognosis; blood viscosity.

3
 ACCEPTED MANUSCRIPT

1. Introduction

Stroke is the second leading cause of death and the most common cause of long-term

disability in adults[1]. Hypoxia often happens during the first days after acute

ischemic stroke onset and it has been reported to be able to result in secondary brain

PT
damage[2, 3]. In order to increase the blood oxygen, supplemental oxygen is often

RI
used to improve prognosis in clinical treatment, but sometimes it may be harmful to

SC
patients[3]. Hemoglobin, a well-known iron-containing oxygen-transport protein[4],

is commonly used in clinical practice as a marker of anemia (low hemoglobin) and

NU
vascular blood clotting (high hemoglobin)[5]. Previous studies demonstrated that
MA

hemoglobin concentrations could affect cardiovascular system through oxygen supply,

blood viscosity and vasoconstriction[6, 7]. Low hemoglobin levels have been shown
ED

to be able to increase the risk of cardiovascular death[8], and elevated hemoglobin

T

levels are also associated with the risk of hypertension and carotid
EP

atherosclerosis[9-11]. In term of stroke, both lower and higher extremes of

hemoglobin concentration were found to be positively associated with the incidence

C
AC

of ischemic stroke[12]. However, few studies investigating the association between

hemoglobin and prognosis of ischemic stroke have yielded inconsistent results[13, 14],

which may be due to small size of sample, incomplete data or heterogeneous subjects

simultaneously including ischemic stroke patients with thrombolytic therapy and

those without thrombolytic therapy. Therefore, large-sample and well-designed

prospective studies on the relationship between hemoglobin levels and clinical

4
 ACCEPTED MANUSCRIPT

outcomes after ischemic stroke are needed. Given that high levels of hemoglobin can

increase blood viscosity which may result in worse cardiovascular function, we

hypothesize that elevated hemoglobin levels in the acute phase are associated with

poor prognosis at 3 months after ischemic stroke. This study prospectively examined

whether the hemoglobin level was independently associated with prognosis of

PT
ischemic stroke based on China Antihypertensive Trial in Acute Ischemic Stroke

RI
(CATIS).

2. Methods
SC
NU
2.1 Study patients
MA

The present study was conducted among ischemic stroke patients from the CATIS

study, a multicenter, single-blind, blinded end-points randomized clinical trial across

ED

China. Details of trial design, methods and major findings of the study have been
T

described previously[15]. Briefly, 4071 ischemic stroke patients from 26 hospitals

EP

were recruited between August 2009 and May 2013 to examine whether immediate

blood pressure (BP) reduction in patients with acute ischemic stroke would reduce
C
AC

death and major disability at 14 days or hospital discharge. The inclusion criteria of

CATIS were to meet all of the followings: age of 22 years or older, having ischemic

stroke confirmed by computed tomography or magnetic resonance imaging of the

brain within 48 hours of symptom onset, and having an elevated systolic BP between

140 mm Hg and less than 220 mm Hg. Patients with a BP ≥ 220/120 mm Hg, severe

heart failure, acute myocardial infarction, unstable angina, atrial fibrillation, aortic

5
 ACCEPTED MANUSCRIPT

dissection, cerebrovascular stenosis (≥70%), resistant hypertension, deep coma or

treatment of intravenous thrombolytic therapy were excluded from the CATIS. In this

study, 190 patients were further excluded because we lacked their baseline records of

hemoglobin levels (n=94) or they were lost to follow-up at 3 months (n=96). Finally, a

total of 3881 acute ischemic stroke patients were included in present analysis (Figure

PT
1).

RI
This study was approved by the institutional review boards at Soochow University in

SC
China and Tulane University in the United States, as well as ethical committees at the

participating hospitals. All study participants or their immediate family members

NU
agreed to take part in the study and provided written informed consents.
MA

2.2 Hemoglobin measurement

ED

Blood samples were collected after at least 8 h of fasting within 24 h of hospital

T

admission. Hemoglobin concentrations were measured for all enrolled patients in each
EP

participating hospital at admission.

C
AC

2.3 Outcome assessment

The primary outcome of this study was a combination of death and major disability

(modified Rankin Scale [mRS] score, 3-6) at 3 months after stroke onset[15-17]. The

secondary outcomes separately those of major disability (mRS, 3-5) and death (mRS,

6). Death certificates were obtained for deceased patients. A trial-wide outcomes

assessment committee, blinded to treatment assignment, reviewed and adjudicated

6
 ACCEPTED MANUSCRIPT

subsequent outcomes based on the criteria established in the Antihypertensive and

Lipid-Lowering Treatment to Prevent Heart Attack Trial.

2.4 Assessment of potential covariates

Demographic characteristics (age, sex, current cigarette smoking, current alcohol

PT
drinking), medical history, the time from onset to randomization and anthropometric

RI
indicators were collected at the time of enrollment[15]. According to the symptoms

SC
and imaging data of the patients, ischemic stroke were classified as large artery

atherosclerosis (thrombotic), cardiac embolism (embolic) and small artery occlusion

NU
lacunae (lacunar) [18]. Stroke severity was estimated by trained neurologists at
MA

baseline using the National Institutes of Health Stroke Scale (NIHSS)[19]. Systolic

and diastolic BP was defined as the average of three measurements when the patient
ED

was in the supine position using a standard mercury sphygmomanometer. Other

T

clinical features including routine laboratory analyses (blood lipids, fasting plasma
EP

glucose, etc.) were performed at admission for all enrolled patients in each

participating hospital.
C
AC

2.5 Statistical analysis

Hemoglobin at baseline was categorized into four groups: Q1 (< 131 g/L), Q2

(131-142 g/L), Q3 (142-152 g/L) and Q4 (≥152 g/L) according to hemoglobin

quartiles. Tests for linear trend of baseline characteristics across hemoglobin quartiles

were performed using covariance analysis for continuous variables and chi-square

7
 ACCEPTED MANUSCRIPT

trend analysis for categorical variables. Logistic regression analysis was used to

estimate the association between hemoglobin level and poor clinical outcomes by

calculating odds ratio (OR) and 95% confidence interval (CI) for higher quartiles

compared to the lowest quartile. Trends for the ORs of ischemic stroke prognosis

across increasing hemoglobin categories were determined, having hemoglobin

PT
category as an ordinal variable. We performed two multiple-adjusted logistic

RI
regression models. In the Model 1, the covariates included age, sex, time from onset

SC
to hospitalization, antihypertensive treatment, current smoking, alcohol consumption,

body mass index, dyslipidemia, fasting plasma glucose and SBP at baseline, ischemic
NU
stroke subtypes, history of hypertension, diabetes and coronary heart disease, and
MA

family history of stroke. In the Model 2, we adjusted for the factors in Model 1 and

further adjusted for baseline NIHSS score. In addition, spline regression models were
ED

used to test linearity assumption of association between hemoglobin and clinical

T

outcomes, fitting a restricted cubic spline function with four knots (at the 5th, 35th,
EP

65th, and 95th percentiles)[20]. We further assessed the potential effect modification

by some stroke-related risk factors (i.e., age, sex, education, BMI, admission NIHSS
C
AC

score, current cigarette smoking, current alcohol drinking, history of hypertension,

receiving immediate BP reduction). Interactions between hemoglobin and subgroup

variables on the primary outcome were tested in the models with interaction terms by

the likelihood ratio test, adjusting for the aforementioned covariates unless the

variable was used as a subgroup variable. All P values were 2-tailed, and a

significance level of 0.05 was used. Statistical analysis was conducted using SAS

8
 ACCEPTED MANUSCRIPT

statistical software (version 9.4, Cary, North Carolina, USA).

3. Result

3.1 Baseline characteristics

A total of 3883 patients (2488 men and 1395 women; mean age, 62.5 ± 10.9 years)

PT
were included in the present study. The median hemoglobin concentration was 142.0

RI
g/L (interquartile range, 131.0-152.0 g/L). The patients with higher hemoglobin were

SC
more likely to be younger, male, cigarette smokers, alcohol drinkers; to have higher

admission diastolic BP, BMI, and fasting plasma glucose; to have a higher prevalence
NU
of dyslipidemia, history of hyperlipidemia and family history of stroke; to have lower
MA

NIHSS score; and to have lower prevalence of coronary heart disease history,

compared with those with lower hemoglobin (Table 1).

T ED

3.2 Hemoglobin level and clinical outcomes

EP

At 3-month follow-up visit, 979 (25.2%) patients experienced the primary outcome

(860 major disabilities and 119 deaths). After adjustment for age, sex and other
C
AC

potential confounders in model 1, the adjusted OR of primary outcome associated

with the highest quartile of hemoglobin was 1.34 (95% CI, 1.04-1.74; Ptrend = 0.021).

We further adjusted for baseline NIHSS score in model 2 and the adjusted OR of

primary outcome for highest vs. lowest quartile of hemoglobin was 1.38 (95%CI,

1.03-1.86; Ptrend = 0.043). Similar significant findings were observed for major

disability in model 1 (OR: 1.44; 95%CI, 1.11-1.89; Ptrend = 0.004) and model 2 (OR:

9
 ACCEPTED MANUSCRIPT

1.49; 95%CI, 1.11-1.99; Ptrend = 0.006), respectively. However, there was no

significant association of baseline hemoglobin with death within 3 months after stroke

(Table 2).

We further used multivariable spline regression models to test linearity assumption of

association between hemoglobin levels and 3-month clinical outcomes after stroke

PT
onset. As shown in Figure 2A, there was a linear relationship between hemoglobin

RI
levels and 3-month primary outcome (P for linearity =0.037). In addition, similar

SC
linear association of hemoglobin levels with major disability was observed among

ischemic stroke patients (P for linearity =0.004, Figure 2B).

NU
In subgroup analyses stratified by age, sex, education, BMI, admission NIHSS score,
MA

current cigarette smoking, current alcohol drinking, history of hypertension, receiving

immediate BP reduction, the modest positive associations of hemoglobin with primary

ED

outcome and major disability were observed in almost all subgroups, and reached
T

statistical significance in several subgroups. Moreover, no significant interaction was

EP

observed between hemoglobin levels and these interested factors in relation to

primary outcome or major disability among ischemic stroke patients (Pinteraction >0.05
C
AC

for all, Figure 3).

4. Discussion

In this prospective study based on CATIS, elevated hemoglobin levels at baseline

were associated with 3-month poor prognosis after adjustment for conventional

prognostic factors among 3883 ischemic stroke patients. The associations between

10
 ACCEPTED MANUSCRIPT

them remained significant after further adjustment for NIHSS. Moreover, a linear

association was observed between baseline hemoglobin levels and 3-month poor

prognosis after ischemic stroke onset. These results suggested that hemoglobin might

be a potential biomarker to predict clinical outcomes at 3 months after ischemic

stroke.

PT
The findings of present study may have several important clinical implications. The

RI
latest guidelines for the early management of acute ischemic stroke patients from

SC
American Heart Association/American Stroke Association (AHA/ASA) have

mentioned several recommendations of general supportive care and emergency

NU
treatment, which includes supplemental oxygen, volume expansion/hemodilution and
MA

anticoagulants[21]. High levels of hemoglobin are known to increase blood viscosity,

which may in turn increase BP and worsen cardiovascular function[22, 23]. In

ED

addition, it has been suggested that hemoglobin concentrations are significantly higher
T

in individuals with hypertension than in those without hypertension[9, 22, 24-26]. In

EP

the present study, high hemoglobin levels at baseline were associated with poor

prognosis of ischemic stroke after adjusting for several important confounders,

C
AC

indicating that hemoglobin can provide additional prognostic information based on

the established prognostic factors for ischemic stroke patients.

As a routine laboratory index in clinical practice, it is simple and easy to test

hemoglobin levels at acute phase. Therefore, our findings suggest that hemoglobin

levels should be evaluated at admission, and patients with high hemoglobin level

should be aggressively monitored although the guidelines have not yet recommended

11
 ACCEPTED MANUSCRIPT

the ways to deal with elevated hemoglobin.

In fact, high hemoglobin levels have been reported to be positively associated with the

risk of various cardiovascular diseases, including hypertension, atherosclerosis and

ischemic stroke[5, 9, 12, 25]. However, reports regarding relationship between

hemoglobin levels and prognosis of ischemic stroke are inconsistent so far. Kellert et

PT
al[27] reported that low hemoglobin was associated with poor clinical outcome among

RI
236 ischemic stroke patients. Barlas et al[14] reported a U-shaped curve relationship

SC
between hemoglobin and mortality after stroke among 6951 patients from a

hospital-based register database, in which extreme low and extreme high hemoglobin
NU
levels were both associated with increased risk of mortality after ischemic stroke.
MA

However, the participants in that study is heterogeneous because ischemic stroke

patients with thrombolytic therapy and those without thrombolytic therapy are all
ED

included. Our study is a multicenter prospective study with a relatively large sample
T

size of acute ischemic stroke patients with elevated systolic BP. In our study, we were
EP

not able to analyze the relationship between extreme low hemoglobin and poor

prognosis after ischemic stroke because of rare patients with extreme low hemoglobin.
C
AC

However, such characteristics that all participants had elevated systolic BP might

enable us to provide more valid appraisal of relationship between high hemoglobin

and poor prognosis after ischemic stroke. The present study had rigid quality control

procedures for baseline data collection and outcome assessment. Furthermore,

comprehensive information about potential confounders were collected and adjusted

in the multivariate models. In addition, the participants treated with intravenous

12
 ACCEPTED MANUSCRIPT

thrombolytic therapy which might increase hemoglobin concentrations[28] were

excluded, so the hemoglobin levels at baseline were not affected by thrombolytic

therapy. We found that elevated hemoglobin was associated with poor outcomes at 3

months among 3883 ischemic stroke patients with an elevated systolic BP. Moreover,

elevated hemoglobin remained independently associated with poor prognosis even

PT
after adjusting for baseline NIHSS score. We also found that the prognostic value of

RI
hemoglobin seems to be stronger for major disability than for death, indicating that

SC
elevated hemoglobin may mainly affect functional recovery of ischemic stroke

patients. In addition, rates of slight stroke in CATIS is relatively high [median NIHSS
NU
score: 4 (2-7)]. Therefore, relatively few numbers of death were observed during
MA

3-month follow-up, which might limit our power to detect significant association

between hemoglobin and death. Further long-term follow-up studies are required to
ED

examine the association between them.

T

Although the mechanisms underlying association between increased hemoglobin and

EP

poor prognosis after ischemic stroke are not clear yet, several potential

pathophysiological pathways have been suggested. Elevated BP is common in the

C
AC

acute phase of ischemic stroke[29], which can lead to the activation of the

renin-angiotensin-aldosterone system and then result in angiotensin-2 producing,

vasoconstriction and erythropoietin production[30]. In addition, endothelial cell

damage may increase not only BP but also growth factor level[31],which enhances

hematopoiesis and then increases hemoglobin level[32].In addition, both increased

blood viscosity[33] and iron overload[34] due to high hemoglobin can affect coronary,

13
 ACCEPTED MANUSCRIPT

cerebral, as well as peripheral blood flow and perfusion[35-37]. Blood rheology is

important in the brain microcirculation, and even a small reduction in blood flow may

have a huge impact on cerebral function, especially in acute phase of ischemic

stroke[38]. Elevated hemoglobin can increase blood viscosity, and further increase

peripheral resistance and diminished cardiac output[5]. High hemoglobin levels may

PT
also aggregate erythrocyte, and then lead to platelet aggregation and adhesion on the

RI
arterial wall[33, 37, 39].

SC
Our study has some limitations. First, this study was not applicable to assess the

association of low hemoglobin levels with poor outcomes after ischemic stroke
NU
because of a low rate of anemia among the participants. Second, the selection bias
MA

may exist in the present study, but the baseline characteristics of participants are

similar to those from the China National Stroke Registry, suggesting that the selection
ED

bias may be minimal[40]. Third, we did not collect the data of other high
T

hemoglobin-related factors such as iron intake, the altitude of residence, and lung
EP

function/disease, so there may be a possibility of residual confounding in this study.

C
AC

5. Conclusions

In summary, elevated hemoglobin levels in the acute phase were associated with poor

prognosis at 3 months after ischemic stroke. Further prospective studies from other

samples of ischemic stroke patients are needed to validate our findings.

Sources of funding

14
 ACCEPTED MANUSCRIPT

This study was supported by the National Natural Science Foundation of China

(grants: 81673263 and 81320108026) and a Project of the Priority Academic Program

Development of Jiangsu Higher Education Institutions, China; Tulane University and

Collins C. Diboll Private Foundation, both in New Orleans, Louisiana.

PT
Disclosures

RI
None.

Author contributions
SC
NU
Yonghong Zhang and Jiang He conceived and designed the study. Yanbo Peng, Tan
MA

Xu, and Yonghong Zhang coordinated the study. Daoxia Guo, Zhengbao Zhu,

Chongke Zhong, Hao Peng, Tian Xu, Aili Wang, Yanbo Peng, Tan Xu, Chung-Shiuan
ED

Chen, Yongqiu Li, Zhong Ju, Jing Chen, Yonghong Zhang, and Jiang He oversaw
T

subject recruitment and monitored gathering of clinical data. Daoxia Guo and
EP

Zhengbao Zhu conducted the statistical analysis and prepared the paper. Yonghong

Zhang and Jiang He revised the paper.

C
AC

Acknowledgements

We thank the study participants and their relatives and the clinical staff at all

participating hospitals for their support and contribution to this project.

Reference

15
 ACCEPTED MANUSCRIPT

[1] G.A. Donnan, M. Fisher, M. Macleod, S.M. Davis, Stroke, Lancet 371(9624)

(2008) 1612-23.

[2] C. Roffe, S. Sills, M. Halim, K. Wilde, M.B. Allen, P.W. Jones, P. Crome,

Unexpected nocturnal hypoxia in patients with acute stroke, S troke 34(11) (2003)

2641-5.

PT
[3] B.R. O'Driscoll, L.S. Howard, J. Earis, V. Mak, G. British Thoracic Society

RI
Emergency Oxygen Guideline, B.T.S.E.O.G.D. Group, BTS guideline for oxygen use

SC
in adults in healthcare and emergency settings, Thorax 72(Suppl 1) (2017) ii1-ii90.

[4] A. D'Alessandro, P.G. Righetti, L. Zolla, The red blood cell proteome and
NU
interactome: an update, Journal of proteome research 9(1) (2010) 144-63.
MA

[5] I. Holme, A.H. Aastveit, N. Hammar, I. Jungner, G. Walldius, High blood

hemoglobin concentration as risk factor of major atherosclerotic cardiovascular events

ED

in 114,159 healthy men and women in the apolipoprotein mortality risk study
T

(AMORIS), Annals of medicine 44(5) (2012) 476-86.

EP

[6] P. Cabrales, G. Han, P. Nacharaju, A.J. Friedman, J.M. Friedman, Reversal of

hemoglobin- induced vasoconstriction with sustained release of nitric oxide, American

C
AC

journal of physiology. Heart and circulatory physiology 300(1) (2011) H49-56.

[7] P. Cabrales, G. Sun, Y. Zhou, D.R. Harris, A.G. Tsai, M. Intaglietta, A.F. Palmer,

Effects of the molecular mass of tense-state polymerized bovine hemoglobin on blood

pressure and vasoconstriction, Journal of applied physiology 107(5) (2009) 1548-58.

[8] P.R. Kalra, N. Greenlaw, R. Ferrari, I. Ford, J.C. Tardif, M. Tendera, C.M. Reid, N.

Danchin, J. Stepinska, P.G. Steg, K.M. Fox, F.p.w.s.c.a.d.I. ProspeCtive observational

16
 ACCEPTED MANUSCRIPT

LongitudinAl RegIstry o, Hemoglobin and Change in Hemoglobin Status Predict

Mortality, Cardiovascular Events, and Bleeding in Stable Coronary Artery Disease,

The American journal of medicine 130(6) (2017) 720-730.

[9] F. Atsma, I. Veldhuizen, W. de Kort, M. van Kraaij, P. Pasker-de Jong, J. Deinum,

Hemoglobin level is positively associated with blood pressure in a large cohort of

PT
healthy individuals, Hypertension 60(4) (2012) 936-41.

RI
[10] Y. Shimizu, M. Nakazato, T. Sekita, K. Kadota, K. Arima, H. Yamasaki, N.

SC
Takamura, K. Aoyagi, T. Maeda, Association between the hemoglobin levels and

hypertension in relation to the BMI status in a rural Japanese population: the Nagasaki
NU
Islands Study, Internal medicine 53(5) (2014) 435-40.
MA

[11] C. Irace, M. Ciamei, A. Crivaro, E. Fiaschi, A. Madia, C. Cortese, A. Gnasso,

Hematocrit is associated with carotid atherosclerosis in men but not in women,

ED

Coronary artery disease 14(4) (2003) 279-84.

T

[12] B. Panwar, S.E. Judd, D.G. Warnock, W.M. McClellan, J.N. Booth, 3rd, P.
EP

Muntner, O.M. Gutierrez, Hemoglobin Concentration and Risk of Incident Stroke in

Community- Living Adults, Stroke 47(8) (2016) 2017-24.

C
AC

[13] Y.H. Park, B.J. Kim, J.S. Kim, M.H. Yang, M.S. Jang, N. Kim, M.K. Han, J.S.

Lee, J. Lee, S. Kim, H.J. Bae, Impact of both ends of the hemoglobin range on clinical

outcomes in acute ischemic stroke, Stroke 44(11) (2013) 3220-2.

[14] R.S. Barlas, K. Honney, Y.K. Loke, S.J. McCall, J.H. Bettencourt-Silva, A.B.

Clark, K.M. Bowles, A.K. Metcalf, M.A. Mamas, J.F. Potter, P.K. Myint, Impact of

Hemoglobin Levels and Anemia on Mortality in Acute Stroke: Analysis of UK

17
 ACCEPTED MANUSCRIPT

Regional Registry Data, Systematic Review, and Meta-Analysis, Journal of the

American Heart Association 5(8) (2016) pii: e003019.

[15] J. He, Y. Zhang, T. Xu, Q. Zhao, D. Wang, C.S. Chen, W. Tong, C. Liu, T. Xu, Z.

Ju, Y. Peng, H. Peng, Q. Li, D. Geng, J. Zhang, D. Li, F. Zhang, L. Guo, Y. Sun, X.

Wang, Y. Cui, Y. Li, D. Ma, G. Yang, Y. Gao, X. Yuan, L.A. Bazzano, J. Chen, C.

PT
Investigators, Effects of immediate blood pressure reduction on death and major

RI
disability in patients with acute ischemic stroke: the CATIS randomized clinical trial,

SC
Jama 311(5) (2014) 479-89.

[16] C.S. Anderson, T. Robinson, R.I. Lindley, H. Arima, P.M. Lavados, T.H. Lee, J.P.
NU
Broderick, X. Chen, G. Chen, V.K. Sharma, J.S. Kim, N.H. Thang, Y. Cao, M.W.
MA

Parsons, C. Levi, Y. Huang, V.V. Olavarria, A.M. Demchuk, P.M. Bath, G.A. Donnan,

S. Martins, O.M. Pontes-Neto, F. Silva, S. Ricci, C. Roffe, J. Pandian, L. Billot, M.

ED

Woodward, Q. Li, X. Wang, J. Wang, J. Chalmers, E. Investigators, Coordinators,

T

Low-Dose versus Standard-Dose Intravenous Alteplase in Acute Ischemic Stroke, The

EP

New England journal of medicine 374(24) (2016) 2313-23.

[17] C.S. Anderson, E. Heeley, Y. Huang, J. Wang, C. Stapf, C. Delcourt, R. Lindley,

C
AC

T. Robinson, P. Lavados, B. Neal, J. Hata, H. Arima, M. Parsons, Y. Li, J. Wang, S.

Heritier, Q. Li, M. Woodward, R.J. Simes, S.M. Davis, J. Chalmers, I. Investigators,

Rapid blood-pressure lowering in patients with acute intracerebral hemorrhage, The

New England journal of medicine 368(25) (2013) 2355-65.

[18] H.P. Adams, Jr., B.H. Bendixen, L.J. Kappelle, J. Biller, B.B. Love, D.L. Gordon,

E.E. Marsh, 3rd, Classification of subtype of acute ischemic stroke. Definitions for

18
 ACCEPTED MANUSCRIPT

use in a multicenter clinical trial. TOAST. Trial of Org 10172 in Acute Stroke

Treatment, Stroke 24(1) (1993) 35-41.

[19] T. Brott, H.P. Adams, Jr., C.P. Olinger, J.R. Marler, W.G. Barsan, J. Biller, J.

Spilker, R. Holleran, R. Eberle, V. Hertzberg, et al., Measurements of acute cerebral

infarction: a clinical examination scale, Stroke 20(7) (1989) 864-70.

PT
[20] S. Durrleman, R. Simon, Flexible regression models with cubic splines, Statistics

RI
in medicine 8(5) (1989) 551-61.

SC
[21] W.J. Powers, A.A. Rabinstein, T. Ackerson, O.M. Adeoye, N.C. Bambakidis, K.

Becker, J. Biller, M. Brown, B.M. Demaerschalk, B. Hoh, E.C. Jauch, C.S. Kidwell,
NU
T.M. Leslie-Mazwi, B. Ovbiagele, P.A. Scott, K.N. Sheth, A.M. Southerland, D.V.
MA

Summers, D.L. Tirschwell, C. American Heart Association Stroke, 2018 Guidelines

for the Early Management of Patients With Acute Ischemic Stroke: A Guideline for
ED

Healthcare Professionals From the American Heart Association/American Stroke

T

Association, Stroke 49(3) (2018) e46-e110.

EP

[22] B.O. Gobel, A. Schulte-Gobel, B. Weisser, K. Glanzer, H. Vetter, R. Dusing,

Arterial blood pressure. Correlation with erythrocyte count, hematocrit, and

C
AC

hemoglobin concentration, American journal of hypertension 4(1 Pt 1) (1991) 14-9.

[23] G.D. Lowe, A.J. Lee, A. Rumley, J.F. Price, F.G. Fowkes, Blood viscosity and

risk of cardiovascular events: the Edinburgh Artery Study, British journal of

haematology 96(1) (1997) 168-73.

[24] N.H. Kim, J.M. Lee, H.C. Kim, J.Y. Lee, H. Yeom, J.H. Lee, I. Suh,

Cross-sectional and longitudinal association between hemoglobin concentration and

19
 ACCEPTED MANUSCRIPT

hypertension: A population-based cohort study, Medicine 95(41) (2016) e5041.

[25] L. Ren, B. Gu, Y. Du, X. Wu, X. Liu, H. Wang, L. Jiang, Y. Guo, J. Wang,

Hemoglobin in normal range, the lower the better?-Evidence from a study from

Chinese community-dwelling participants, Journal of thoracic disease 6(5) (2014)

477-82.

PT
[26] S.G. Lee, J.H. Rim, J.H. Kim, Association of hemoglobin levels with blood

RI
pressure and hypertension in a large population-based study: the Korea National

SC
Health and Nutrition Examination Surveys 2008-2011, Clinica chimica acta;

international journal of clinical chemistry 438 (2015) 12-8.

NU
[27] L. Kellert, E. Martin, M. Sykora, H. Bauer, P. Gussmann, J. Diedler, C. Herweh,
MA

P.A. Ringleb, W. Hacke, T. Steiner, J. Bosel, Cerebral oxygen transport failure?:

decreasing hemoglobin and hematocrit levels after ischemic stroke predict poor
ED

outcome and mortality: STroke: RelevAnt Impact of hemoGlobin, Hematocrit and

T

Transfusion (STRAIGHT)--an observational study, Stroke 42(10) (2011) 2832-7.

EP

[28] J.T. Sertorio, E.M. Neto-Neves, C.A. Dias-Junior, O. Sousa-Santos, T. Kiss, D.

Muhl, J.E. Tanus-Santos, Elevated plasma hemoglobin levels increase nitric oxide
C
AC

consumption in experimental and clinical acute pulmonary thromboembolism, Critical

care medicine 41(7) (2013) e118-24.

[29] A.I. Qureshi, M.A. Ezzeddine, A. Nasar, M.F. Suri, J.F. Kirmani, H.M. Hussein,

A.A. Divani, A.S. Reddi, Prevalence of elevated blood pressure in 563,704 adult

patients with stroke presenting to the ED in the United States, The American journal

of emergency medicine 25(1) (2007) 32-8.

20
 ACCEPTED MANUSCRIPT

[30] S.M. Freudenthaler, T. Schenck, I. Lucht, C.H. Gleiter, Fenoterol stimulates

human erythropoietin production via activation of the renin angiotensin system,

British journal of clinical pharmacology 48(4) (1999) 631-4.

[31] Y. Nakamura, R. Morishita, J. Higaki, I. Kida, M. Aoki, A. Moriguchi, K.

Yamada, S. Hayashi, Y. Yo, H. Nakano, K. Matsumoto, T. Nakamura, T. Ogihara,

PT
Hepatocyte growth factor is a novel member of the endothelium- specific growth

RI
factors: additive stimulatory effect of hepatocyte growth factor with basic fibroblast

SC
growth factor but not with vascular endothelial growth factor, Journal of hypertension

14(9) (1996) 1067-72.

NU
[32] K. Takai, J. Hara, K. Matsumoto, G. Hosoi, Y. Osugi, A. Tawa, S. Okada, T.
MA

Nakamura, Hepatocyte growth factor is constitutively produced by human bone

marrow stromal cells and indirectly promotes hematopoiesis, Blood 89(5) (1997)
ED

1560-5.
T

[33] J.H. Wood, D.B. Kee, Jr., Hemorheology of the cerebral circulation in stroke,
EP

Stroke 16(5) (1985) 765-72.

[34] D. Pratico, M. Pasin, O.P. Barry, A. Ghiselli, G. Sabatino, L. Iuliano, G.A.

C
AC

FitzGerald, F. Violi, Iron-dependent human platelet activation and hydroxyl radical

formation: involvement of protein kinase C, Circulation 99(24) (1999) 3118-24.

[35] G.D. Lowe, C.D. Forbes, Blood rheology and thrombosis, Clinics in

haematology 10(2) (1981) 343-67.

[36] J.A. Dormandy, E. Hoare, J. Colley, D.E. Arrowsmith, T.L. Dormandy, Clinical,

haemodynamic, rheological, and biochemical findings in 126 patients with

21
 ACCEPTED MANUSCRIPT

intermittent claudication, British medical journal 4(5892) (1973) 576-81.

[37] J.A. Dormandy, E. Hoare, A.H. Khattab, D.E. Arrowsmith, T.L. Dormandy,

Prognostic significance of rheological and biochemical findings in patients with

intermittent claudication, British medical journal 4(5892) (1973) 581-3.

[38] R. Muller, F. Lehrach, Haemorheology and cerebrovascular disease:

PT
multifunctional approach with pentoxifylline, Current medical research and opinion

RI
7(4) (1981) 253-63.

SC
[39] D.J. Thomas, J. Marshall, R.W. Russell, G. Wetherley-Mein, G.H. du Boulay, T.C.

Pearson, L. Symon, E. Zilkha, Effect of haematocrit on cerebral blood-flow in man,

NU
Lancet 2(8045) (1977) 941-3.
MA

[40] Y. Luo, X. Wang, K. Matsushita, C. Wang, X. Zhao, B. Hu, L. Liu, H. Li, G. Liu,

Q. Jia, Y. Wang, Y. Wang, C. Investigators, Associations between estimated

ED

glomerular filtration rate and stroke outcomes in diabetic versus nondiabetic patients,
T

Stroke 45(10) (2014) 2887-93.

C EP
AC

22
 ACCEPTED MANUSCRIPT

Figure captions

Figure 1. Study participants flow chart.

Figure 2. Linear test of the association between baseline blood hemoglobin and

3-month clinical outcomes.

Odds ratio and 95% confidence interval derived from restricted cubic spline

PT
regression, with knots placed at the 5th, 35th, 65th, and 95th percentiles of

RI
hemoglobin. Panels adjusted for the same variables as model 2 in Table 2. (A)

SC
Primary outcome; (B) Major disability. There was a linear association between

hemoglobin levels and 3-month poor prognosis of ischemic stroke (primary outcome:
NU
P for linearity =0.037; major disability: P for linearity =0.004).
MA

Figure 3. Subgroup analyses of the association between baseline blood hemoglobin

and clinical outcomes of ischemic stroke at 3 months.

ED

In the multivariate models, confounding factors such as age, sex, time from onset to
T

hospitalization, antihypertensive treatment, current smoking, alcohol consumption,

EP

body mass index, dyslipidemia, fasting plasma glucose, systolic BP and NIHSS score

at baseline, ischemic stroke subtypes, history of hypertension, history of diabetes,

C
AC

history of coronary heart disease and family history of stroke were included unless the

variable was used as a subgroup variable.

23
 ACCEPTED MANUSCRIPT

Highlights

•Elevated hemoglobin was associated with poor outcomes at 3 months after stroke

onset.

•This study included 3883 ischemic stroke patients with an elevated SBP.

•Patients with high hemoglobin level should be aggressively monitored.

PT
•Hemoglobin is simple and easy to test and evaluate at acute phase or admission.

RI
SC
NU
MA
T ED
C EP
AC

24
 ACCEPTED MANUSCRIPT

Hemoglobin level and clinical outcomes at three months after ischemic stroke

Tables

PT
RI
SC
NU
MA
T ED
C EP
AC

25
 ACCEPTED MANUSCRIPT

Table 1. Characteristics of participants according to hemoglobin quartile.

hemoglobin, g/L

Characteristics Q1 (< Q2 Q3 Q4 P trend

Total
131) (131-142) (142-152) (≥152)

Number of subjects 3883 910 990 986 997

PT
Demographic

RI
62.5 ± 66.6 ± 64.4 ± 61.5 ± 57.8 ±
Age, years <0.001

SC
10.9 11.2 9.9 10.2 10.5

2488 288 548 739 913

NU
Male sex <0.001
(64.1) (31.7) (55.4) (75.0) (91.6)
MA

Current cigarette 1440 171 334 413 522

<0.001
smoking (37.1) (18.8) (33.7) (41.9) (52.4)
ED

Current alcohol 1212 110 253 362 487

<0.001
T

drinking (31.2) (12.1) (25.6) (36.7) (48.9)

EP

Clinical features
C

10.0 12.0 10.3

Time from onset to 9.8 (4.0, 10.0 (4.7,
AC

(4.5, (5.0, (5.0, 0.689

randomization, h* 24.0) 24.0)
24.0) 24.0) 24.0)

Baseline systolic BP, 166.1 ± 166.2 ± 165.6 ± 166.2 ± 166.5 ±

0.534
mm Hg 16.9 16.7 16.5 17.2 17.2

Baseline diastolic BP, 96.7 ± 93.9 ± 95.7 ± 97.3 ± 99.5 ±

<0.001
mm Hg 11.1 10.9 10.3 11.1 11.4
26
 ACCEPTED MANUSCRIPT

25 ± 24.7 ± 24.9 ± 25.1 ± 25.3 ±

2
Body mass index, kg/m <0.001
3.1 3.4 3.3 2.9 2.9

2119 476 512 548 583

Dyslipidemia 0.002
(54.6) (52.3) (51.7) (55.6) (58.5)

4.0 5.0 4.0

PT
4.0 (2.0, 4.0 (2.0,
Baseline NIHSS score* (2.0, (3.0, (2.0, 0.017
7.0) 7.0)

RI
7.0) 8.0) 7.0)

SC
Baseline fasting plasma 6.7 ± 6.5 ± 6.9 ±
<0.001
glucose, mmol/L 2.8 2.6 6.6 ± 2.6 6.8 ± 2.9 2.9
NU
Medical history
MA

3058 708 781 783 786

History of hypertension 0.553
(78.8) (77.8) (78.9) (79.4) (78.8)
ED

History of 268
0.022
T

hyperlipidemia (6.9) 49 (5.4) 70 (7.1) 66 (6.7) 83 (8.3)

EP

History of diabetes 686 173 166 173 174

0.502
mellitus (17.7) (19) (16.8) (17.6) (17.5)
C
AC

History of coronary 427 127 118

<0.001
heart disease (11) (14) (11.9) 86 (8.7) 96 (9.6)

729 157 176 184 212

Family history of stroke 0.021
(18.8) (17.3) (17.8) (18.7) (21.3)

Ischemic stroke

subtype

27
 ACCEPTED MANUSCRIPT

3028 701 776 774 777

Thrombotic 0.655
(78) (77) (78.4) (78.5) (77.9)

189
Embolic 0.430
(4.9) 48 (5.3) 48 (4.9) 49 (5) 44 (4.4)

765 187 192 190 196

PT
Lacunar 0.641
(19.7) (20.6) (19.4) (19.3) (19.7)

RI
Receiving immediate 1937 462 499 469 507
0.737

SC
BP reduction (49.9) (50.8) (50.4) (47.6) (50.9)

Abbreviations: BP = blood pressure; NIHSS = National Institute of Health Stroke

NU
Scale;
MA

Data were presented as mean ± SD or n (%) unless otherwise noted;

*Data were presented as median (interquartile range).

T ED
C EP
AC

Table 2. Odds ratio and 95% confidence interval of clinical outcomes for quartile of

hemoglobin at baseline.

28
 ACCEPTED MANUSCRIPT

Hemoglobin, g/L

Q1 (< P trend
Q2 (131-142) Q3 (142-152) Q4 (≥152)
131)

Primary outcome: death or major disability (mRS 3-6)

0.93 1.01 1.34

PT
Model 1 ref 0.021
(0.74-1.16) (0.79-1.28) (1.04-1.74)

RI
1.00 1.01 1.38
Model 2 ref 0.043

SC
(0.77-1.29) (0.76-1.33) (1.03-1.86)

Major disability (mRS 3-5)

NU
0.98 1.15 1.44
Model 1 ref 0.004
MA

(0.78-1.25) (0.90-1.48) (1.11-1.89)

1.07 1.20 1.49

Model 2 ref 0.006
ED

(0.83-1.39) (0.92-1.58) (1.11-1.99)

T

Death
EP

0.75 0.44 0.82

Model 1 ref 0.225
C

(0.45-1.26) (0.23-0.84) (0.44-1.51)

AC

0.85 0.43 0.79

Model 2 ref 0.190
(0.49-1.49) (0.21-0.85) (0.41-1.52)

Model1, adjusted for age, sex, time from onset to hospitalization, antihypertensive

treatment, current smoking, alcohol consumption, body mass index, dyslipidemia,

fasting plasma glucose and SBP at baseline, ischemic stroke subtypes, history of

hypertension, diabetes and coronary heart disease, and family history of stroke;
29
 ACCEPTED MANUSCRIPT

Model 2, model 1 plus baseline NIHSS score.

PT
RI
SC
NU
MA
T ED
C EP
AC

30