Yasuteru Inoue, Fumio Miyashita, Kazunori Toyoda and Kazuo MinematsuIntracerebral Hemorrhage

Low Serum Calcium Levels Contribute to Larger Hematoma Volume in Acute

Print ISSN: 0039-2499. Online ISSN: 1524-4628 Copyright © 2013 American Heart Association, Inc. All rights reserved.

is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231Stroke published online May 14, 2013;Stroke. 

http://stroke.ahajournals.org/content/early/2013/05/14/STROKEAHA.113.001187World Wide Web at:

The online version of this article, along with updated information and services, is located on the

http://stroke.ahajournals.org/content/suppl/2013/05/14/STROKEAHA.113.001187.DC1.htmlData Supplement (unedited) at:

  http://stroke.ahajournals.org//subscriptions/

is online at: Stroke Information about subscribing to Subscriptions: 

http://www.lww.com/reprints Information about reprints can be found online at: Reprints:

  document. Permissions and Rights Question and Answer process is available in the

Request Permissions in the middle column of the Web page under Services. Further information about thisOnce the online version of the published article for which permission is being requested is located, click

can be obtained via RightsLink, a service of the Copyright Clearance Center, not the Editorial Office.Strokein Requests for permissions to reproduce figures, tables, or portions of articles originally publishedPermissions:

at WORLD HLTH ORGANIZATION on June 23, 2013http://stroke.ahajournals.org/Downloaded from

1

Intracerebral hemorrhage (ICH) is associated with poor out-come, a high mortality rate, and little effective treatment.1 In

patients with acute ICH, large hematoma volume, the presence of intraventricular bleeding, and prior use of anticoagulants or antiplatelets are reported to be associated with poor outcome.2,3

Several studies have reported that low serum calcium (Ca) levels have an association with large infarcts and poor out-come among patients with ischemic stroke.4–6 In another study involving a majority of patients with ischemic stroke and few patients with ICH, low serum Ca levels were also associated with poor outcome.7 In addition, ionized Ca is an essential cofactor for the coagulation cascade and associated with con-version of prothrombin to thrombin. Association of Ca lev-els with changes in clotting time and bleeding tendency was shown in the rodent models.8,9 Thus, serum Ca levels might play a pivotal role in hemostasis in acute ICH.

However, associations of serum Ca levels with clinical findings and outcomes of isolated patients with ICH remain unknown. Thus, the aim of the present study was to exam-ine the associations between admission Ca levels and clinical findings and outcomes in patients with acute ICH.

Patients and MethodsConsecutive patients admitted to our department within 24 hours from the onset of nontraumatic ICH were studied using our pro-spectively collected database on inpatients with stroke. Patients

were classified into quartiles based on admission serum Ca levels (Q1 [≤9.0], Q2 [9.1–9.3], Q3 [9.4–9.7], Q4 [≥9.8] mg/dL). Acute outcomes included the following: the National Institutes of Health Stroke Scale (NIHSS) score at admission, initial hematoma volume, and hematoma growth. Chronic outcomes included the following: modified Rankin Scale (mRS) scores of 0 to 1 and 0 to 2, bedridden state or death corresponding to mRS scores of 5 to 6 at 30 days, and mortality. For chronic outcomes, patients with prestroke mRS score ≥2 were excluded from the analyses. Associations between each Ca quartile and outcomes were determined using multivariate regression models adjusted by the baseline characteristics automatically selected in a backward stepwise selection method (see Methods in the online-only Data Supplement).

ResultsA total of 273 patients (92 women, 70±11 years old) were studied. There were fewer women (P=0.036), liver dysfunction was more common (P=0.043), and levels of albumin (P<0.001), total cholesterol (P=0.002), low-density lipoprotein–cholesterol (P<0.001), high-density lipoprotein–cholesterol (P=0.027), and hemoglobin (P=0.002) were lower in the lower Ca level quartiles than in the higher Ca level quartiles (see Tables I and II in the online-only Data Supplement).

The lowest Ca quartile had higher hematoma volume (P=0.005; Table 1). After multivariate linear regression analy-ses, Q1 had larger hematoma volume than Q4 (P=0.015). The lowest Ca quartile had a higher NIHSS score (P=0.010).

Brief Report

Background and Purpose—We investigate whether admission serum calcium levels are associated with hematoma volume, stroke severity, and outcomes in patients with acute intracerebral hemorrhage.

Methods—A total of 273 patients admitted within 24 hours after intracerebral hemorrhage onset was divided into quartiles based on admission serum calcium levels (Q1 [≤9.0], Q2 [9.1–9.3], Q3 [9.4–9.7], Q4 [≥9.8] mg/dL).

Results—Median hematoma volumes for each quartile (Q1 to Q4) were 18, 9, 10, and 9 mL (P=0.005), and median National Institutes of Health Stroke Scale scores were 16, 11, 11, and 9 (P=0.010), respectively. On multivariate analysis, Q1 had larger hematoma volume (P=0.025) and higher National Institutes of Health Stroke Scale score (P=0.020) than Q4. There were fewer patients with modified Rankin Scale scores 0 to 2 in Q1 than Q4 after adjustment for risk factors and comorbidities (odds ratio, 0.31; 95% confidence interval, 0.11–0.84) but not after additional adjustment for hematoma volume and National Institutes of Health Stroke Scale score. There were more patients with modified Rankin Scale scores 5 to 6 (P=0.016) and with fatal outcomes (P=0.048) in Q1 than Q4 as crude values, but not after adjustment.

Conclusions—Low admission serum calcium levels were associated with larger hematoma volume and higher National Institutes of Health Stroke Scale score among patients with acute intracerebral hemorrhage. (Stroke. 2013;44:00-00.)

Key Words: acute stroke ■ hematoma ■ intracerebral hemorrhage ■ serum calcium ■ stroke outcome

Low Serum Calcium Levels Contribute to Larger Hematoma Volume in Acute Intracerebral Hemorrhage

Yasuteru Inoue, MD; Fumio Miyashita, MD; Kazunori Toyoda, MD; Kazuo Minematsu, MD

Received February 13, 2013; accepted March 26, 2013.From the Department of Cerebrovascular Medicine, National Cerebral and Cardiovascular Center, Suita, Japan.The online-only Data Supplement is available with this article at http://stroke.ahajournals.org/lookup/suppl/doi:10.1161/STROKEAHA.

113.001187/-/DC1.Correspondence to Kazunori Toyoda, MD, Department of Cerebrovascular Medicine, National Cerebral and Cardiovascular Center, 5-7-1 Fujishiro-dai,

Suita, Osaka 565-8565, Japan. E-mail toyoda@hsp.ncvc.go.jp© 2013 American Heart Association, Inc.

Stroke is available at http://stroke.ahajournals.org DOI: 10.1161/STROKEAHA.113.001187

at WORLD HLTH ORGANIZATION on June 23, 2013http://stroke.ahajournals.org/Downloaded from

2 Stroke July 2013

After multivariate linear regression analyses, Q1 had a higher NIHSS score than Q4 (P=0.010).

The Figure shows the distribution of mRS scores. No patients in Q1 had mRS score 0 to 1, as compared with other quartiles (P<0.001). Patients with mRS score 0 to 2 were less frequent in Q1 than Q4 as crude values (P=0.002) and after adjustment with risk factors and comorbidities (P=0.026), but they were no longer different after additional adjustment with hematoma volume and NIHSS score (Table 2). Patients with mRS score 5 to 6 and those with fatal outcomes were more frequent in Q1 than Q4 as crude values (P=0.016 and 0.048, respectively), but the difference disappeared after multivariate adjustment (Table III in the online-only Data Supplement).

DiscussionThis is the first study that investigated the relationships between serum Ca levels at admission and clinical findings and outcomes of patients with acute ICH. The major new find-ing was that patients with low Ca levels had larger hematoma volumes and higher NIHSS scores at admission.

Three possible mechanisms may explain why serum Ca lev-els are related to severity at admission. First, as stated above, ionized Ca is an essential cofactor for the coagulation cascade. Second, low serum Ca levels might contribute to hematoma

enlargement through blood pressure elevation in the acute ICH. Ca was reported to induce relaxation of isolated arteries by activating Ca receptors in perivascular nerves.10 Third, low serum Ca levels might reflect poor liver function. The low-est Ca quartile (Q1) had high percentage of liver disease, and these findings suggested another mechanism for poor coagula-tion and therefore larger hematoma volume.

This study was limited by its design as a single-center, hospital-based, retrospective analysis with exclusion of some patients attributable to lack of serum Ca level data. To sim-plify reporting of results, data on ionized Ca levels and albu-min-corrected Ca levels were not shown. Instead, the serum albumin level was used for multivariate regression analyses and it did not affect the results.

In conclusion, low Ca levels in the hyperacute stage of ICH are indicators of larger hematoma volume and more severe neurological deficit. There is an apparent association between serum Ca levels and hematoma volume. Future studies could incorporate serum Ca in models and test for enhanced

Table 1. Multivariate Analysis for Hematoma Volume and NIHSS Score

Median Value (Interquartile Ratio) β SE P value

Hematoma volume*

Q1 18 (9–40) 0.26 0.11 0.015

Q2 9 (4–30) 0.16 0.10 0.111

Q3 10 (4–18) 0.03 0.10 0.804

Q4 9 (2–22) 0 (reference)

… …

NIHSS score†

Q1 16 (9–22) 0.18 0.07 0.010

Q2 11 (6–18) 0.07 0.06 0.256

Q3 11 (7–16) 0.10 0.06 0.132

Q4 9 (5–17) 0 (reference)

… …

β: Difference in log-transformed values compared with Q4. NIHSS indicates National Institutes of Health Stroke Scale.

*Adjusted by age, sex, history of stroke, anticoagulant, antiplatelet and statin use, onset-arrival time, intraventricular bleeding, hemoglobin, and prothrombin time-international normalized ratio in a backward stepwise selection method.

†Adjusted by age, sex, anticoagulant, antihypertensive and statin use, onset-arrival time, intraventricular bleeding, phosphorus, aspartate transaminase, glucose, and prothrombin time-international normalized ratio in a backward stepwise selection method.

Figure. Distribution of the modified Rankin Scale (mRS) score at 30 days.

Table 2. Multivariate Analysis for Modified Rankin Scale Score 0 to 2 at 30 Days

Patients, %

Crude Adjusted: Model 1 Adjusted: Model 2

OR (95% CI) P Value OR (95% CI) P Value OR (95% CI) P Value

Q1 8 (14.0) 0.25 (0.09–0.60) 0.002 0.31 (0.11–0.84) 0.026 0.57 (0.17–1.83) 0.353

Q2 28 (37.3) 0.91 (0.45–1.84) 0.785 0.87 (0.38–1.98) 0.732 1.25 (0.46–3.43) 0.667

Q3 21 (31.8) 0.71 (0.34–1.49) 0.363 0.62 (0.26–1.45) 0.276 0.94 (0.35–2.54) 0.897

Q4 23 (39.7) 1 (reference) … 1 (reference) … 1 (reference) …

Patients with prestroke modified Rankin Scale score ≥2 were excluded from the analyses. CI indicates confidence interval; and OR, odds ratio.Model 1: Adjusted by age, sex, anticoagulant and statin use, onset-arrival time, intraventricular bleeding, blood glucose, and prothrombin time-international

normalized ratio.Model 2: Further adjusted by National Institutes of Health Stroke Scale score and hematoma volume.

at WORLD HLTH ORGANIZATION on June 23, 2013http://stroke.ahajournals.org/Downloaded from

Inoue et al Serum Calcium, Hematoma Volume, and Severity in ICH 3

predictive power over existing models. The clinical benefit and therapeutic time window for Ca level modification, as well as the Ca threshold level that is beneficial for ICH pre-vention, remain unknown, and further investigations of these issues will be required.

Sources of FundingThis study was supported in part by grants-in-aid (H23-Junkanki-Ippan-010, H24-Junkanki-Ippan-011) from the Ministry of Health, Labor and Welfare, Japan, and the Intramural Research Fund for Cardiovascular Diseases from the National Cerebral and Cardiovascular Center (H22-4-1, H23-4-3).

DisclosuresNone.

References 1. Qureshi AI, Tuhrim S, Broderick JP, Batjer HH, Hondo H, Hanley

DF. Spontaneous intracerebral hemorrhage. N Engl J Med. 2001;344:1450–1460.

2. Broderick JP, Brott TG, Duldner JE, Tomsick T, Huster G. Volume of intracerebral hemorrhage. A powerful and easy-to-use predictor of 30-day mortality. Stroke. 1993;24:987–993.

3. Toyoda K, Yasaka M, Nagata K, Nagao T, Gotoh J, Sakamoto T, et al. Blood pressure levels and bleeding events during antithrombotic therapy: the Bleeding with Antithrombotic Therapy (BAT) Study. Cerebrovasc Dis. 2009; 27:151–159.

4. Ovbiagele B, Liebeskind DS, Starkman S, Sanossian N, Kim D, Razinia T, et al. Are elevated admission calcium levels associated with better out-comes after ischemic stroke? Neurology. 2006;67:170–173.

5. Buck BH, Liebeskind DS, Saver JL, Bang OY, Starkman S, Ali LK, et al. Association of higher serum calcium levels with smaller infarct volumes in acute ischemic stroke. Arch Neurol. 2007;64:1287–1291.

6. Ovbiagele B, Starkman S, Teal P, Lyden P, Kaste M, Davis SM, et al; VISTA Investigators. Serum calcium as prognosticator in ischemic stroke. Stroke. 2008;39:2231–2236.

7. Appel SA, Molshatzki N, Schwammenthal Y, Merzeliak O, Toashi M, Sela BA, et al. Serum calcium levels and long-term mortality in patients with acute stroke. Cerebrovasc Dis. 2011;31:93–99.

8. Fukuda T, Nakashima Y, Harada M, Toyoshima S, Koshitani O, Kawaguchi Y, et al. Effect of whole blood clotting time in rats with ionized hypocalcemia induced by rapid intravenous citrate infusion. J Toxicol Sci. 2006;31:229–234.

9. Fujita Y, Doi K, Harada D, Kamikawa S. Modulation of physiologi-cal hemostasis by irrigation solution: comparison of various irrigation solutions using a mouse brain surface bleeding model. J Neurosurg. 2010;112:824–828.

10. Bukoski RD, Bian K, Wang Y, Mupanomunda M. Perivascular sensory nerve Ca2+ receptor and Ca2+-induced relaxation of isolated arteries. Hypertension. 1997;30:1431–1439.

at WORLD HLTH ORGANIZATION on June 23, 2013http://stroke.ahajournals.org/Downloaded from

Inoue et al: Serum Calcium, hematoma volume and severity in ICH, p1

ONLINE SUPPLEMENT

Low serum calcium levels contribute to larger hematoma volume in acute intracerebral

hemorrhage

Yasuteru Inoue, et al

Supplemental Methods

Patients and Methods

Consecutive patients admitted to our department within 24 hours from the onset of

nontraumatic ICH from January 2004 through June 2009 were studied using our

prospectively collected database on stroke inpatients. Patients with ICH in association with

vascular malformations, aneurysms, tumors, and impaired coagulation (e.g., disseminated

intravascular coagulation syndrome) and those with primary intraventricular hemorrhage

were excluded from the study. Thus, 383 patients were enrolled. Of these, one with poor

imaging data and 109 with unavailable admission serum Ca data were excluded. Finally, 273

patients (92 women, 70±11 years old) were studied. The baseline characteristics of the 273

patients and the 110 excluded patients were similar, except for a history of liver dysfunction

(10.6% vs. 3.7%; p=0.029) and time from ICH onset to arrival (median 145 min vs.110 min;

p=0.029). The ethics committee approved this study.

CT imaging and image analysis

ICH was verified by emergent computed tomography (CT) immediately after

hospital arrival. Follow-up CT was performed routinely within 24 hours after admission.

Hematoma location and the presence of intraventricular bleeding were recorded on initial CT

scan. Hematoma volumes were measured using the ABC/2 method.1 Intraventricular bleeding

was not included in volume calculations. All CT scans were reviewed and evaluated by two

trained stroke specialists who were blinded to the patients’ clinical status.

Data collection

Baseline clinical and demographic information included age, sex, history of ischemic

stroke or prior ICHs, premorbid modified Rankin Scale (mRS) score, liver dysfunction 1

at WORLD HLTH ORGANIZATION on June 23, 2013http://stroke.ahajournals.org/Downloaded from

Inoue et al: Serum Calcium, hematoma volume and severity in ICH, p2

(cirrhosis or chronic hepatitis on history), hypertension, diabetes mellitus, dyslipidemia,

history of alcohol or tobacco use, and prestroke use of anticoagulants, antiplatelet agents,

antihypertensive agents, and statins. Body mass index and blood pressure (BP) levels on

admission were also documented. The time of ICH onset was defined as the time when the

patient was last known to be normal if the exact time of acute onset of a neurological deficit

could not be clearly specified.

Blood tests were performed at the time of admission before any crystalloid infusion

therapy. Laboratory results, including Ca and the variables listed in Table 2, were recorded,

and the estimated glomerular filtration rate (eGFR) was calculated.2 Enzymatic measurement

of baseline serum Ca levels was performed at a central laboratory with a Hitachi Modular

Analytics analyzer (normal range: 8.6-10.2 mg/dL). Patients were divided into quartiles by

their Ca levels.

Outcomes

As acute outcome measures, the National Institutes of Health Stroke Scale (NIHSS)

score on admission evaluated by stroke experts, initial hematoma volume, and hematoma

growth were assessed. Chronic outcomes were evaluated at 30 days; independence

corresponding to mRS scores of 0-1 and 0-2, bedridden state or death corresponding to mRS

scores of 5-6, and mortality were assessed. For chronic outcomes, 18 patients with prestroke

mRS score ≥2 were excluded from the analyses.

Statistical analysis

Baseline demographics and clinical characteristics were compared across quartiles by serum

Ca levels using the χ2 test for percentages and the Kruskal-Wallis rank-sum test for medians.

Continuous variables are reported as medians (interquartile range) unless stated otherwise.

Fisher’s exact test was used to compare dichotomous variables between groups, and the

Wilcoxon rank-sum test was used for continuous and ordinal variables. The difference in

each quartile of serum Ca levels compared to the highest quartile of Ca levels (Q4) was

measured in relation to hematoma volume, NIHSS score on admission, and outcomes.

Hematoma volume and NIHSS score on admission were log-transformed to approximate 2

at WORLD HLTH ORGANIZATION on June 23, 2013http://stroke.ahajournals.org/Downloaded from

Inoue et al: Serum Calcium, hematoma volume and severity in ICH, p3

normality. Multivariate linear regression analyses for hematoma volume and NIHSS score on

admission and multivariate logistic regression analyses for 30-day outcomes were performed

by adjusting for sex, age, and variables that were automatically selected in a backward

stepwise selection method. A backward selection procedure was performed for each outcome

using p>0.10 of the likelihood ratio test for exclusion. The odds ratio and 95% CI were

obtained. Statistical significance was set at p<0.05. Statistical analysis was performed using

the JMP 8.0 statistical software (SAS Institute Inc, Cary, NC).

Supplemental references

1. Kothari RU, Brott T, Broderick JP, Barsan WG, Sauerbeck LR, Zuccarello M, et al. The

ABCs of measuring intracerebral hemorrhage volumes. Stroke. 1996; 27:1304–1305.

2. Matsuo S, Imai E, Horio M, Yasuda Y, Tomita K, Nitta K, et al. Revised equations for

estimated GFR from serum creatinine in Japan. Am J Kidney Dis. 2009; 53:982–992.

3

at WORLD HLTH ORGANIZATION on June 23, 2013http://stroke.ahajournals.org/Downloaded from

Inoue et al: Serum Calcium, hematoma volume and severity in ICH, p4

Supplemental tables

Table S1. Baseline clinical characteristics according to serum calcium level quartiles

Total Q1 Q2 Q3 Q4

Number of patients 273 62 77 71 63

Age, years 70±11 71±11 69±11 68±11 70±11

Women 92 (33.7) * 17 (27.4) 20 (26.0) 25 (35.2) 30 (47.6)

Comorbidities

Past history of stroke 57 (20.9) 19 (30.6) 15 (19.5) 9 (12.7) 14 (22.2)

Liver dysfunction 29 (10.6) * 12 (19.4) 9 (11.7) 4 (5.6) 4 (6.3)

Prestroke mRS score ≥2 18 (6.6) 8 (12.9) 3 (3.9) 2 (2.8) 5 (7.9)

Risk factors

Hypertension 246 (90.1) 53 (85.5) 71 (92.2) 62 (87.3) 60 (95.2)

Diabetes mellitus 63 (23.1) 14 (22.6) 19 (24.7) 14 (19.7) 16 (25.4)

Dyslipidemia 63 (23.1) 14 (22.6) 19 (24.7) 14 (19.7) 16 (25.4)

Alcohol consumption 152 (55.7) 37 (59.7) 49 (63.6) 36 (50.7) 30 (47.6)

Smoking habit 67 (24.5) 21 (33.9) 16 (20.8) 13 (18.3) 17 (27.0)

Prestroke medications

Anticoagulants 22 (8.1) 6 (9.7) 5 (6.5) 5 (7.0) 6 (9.5)

Antiplatelets 49 (17.9) 13 (21.0) 15 (19.5) 7 (9.9) 14 (22.2)

Antihypertensives 128 (46.9) 29 (46.8) 33 (42.9) 35 (49.3) 31 (49.2)

Statins 28 (10.3) 9 (14.5) 7 (9.1) 6 (8.5) 6 (9.5)

Data are presented as means±SD for age and numbers of patients with percentages in

parentheses for the others. * p<0.05 among quartiles.

4

at WORLD HLTH ORGANIZATION on June 23, 2013http://stroke.ahajournals.org/Downloaded from

Inoue et al: Serum Calcium, hematoma volume and severity in ICH, p5

Table S2. ICH features and data on admission according to serum calcium level quartiles

Total Q1 Q2 Q3 Q4

Number of patients 273 62 77 71 63

ICH features

Time from ICH onset to arrival,

min

145

(70-493)

120

(67-473)

145

(60-601)

180

(75-390)

200

(80-600)

Intraventricular bleeding 99 (36.3) 31 (50.0) 23 (29.9) 23 (32.4) 22 (34.9)

Location of hematoma

Thalamo-ganglionic 193 (70.7) 41 (66.1) 50 (64.9) 55 (77.5) 47 (74.6)

Lobar 44 (16.1) 12 (19.4) 17 (22.1) 8 (11.3) 7 (11.1)

Cerebellar 10 (3.7) 2 (3.2) 3 (3.9) 2 (2.8) 3 (4.8)

Brain stem 26 (9.5) 7 (11.3) 7 (9.1) 6 (8.5) 6 (9.5)

Physiological data on admission

Body mass index 22.4±3.8 21.9±3.8 22.4±4.0 22.9±3.7 21.9±3.8

Systolic blood pressure, mmHg 179.0±30.7 179.2±33.3 179.2±32.9 176.8±28.8 180.9±27.9

Diastolic blood pressure, mmHg 94.4±18.9 91.6±17.4 95.1±21.1 93.6±17.2 97.0±19.2

Blood test results on admission

Calcium, mg/dL 9.4±0.49 * 8.8±0.20 9.2±0.085 9.6±0.10 10.1±0.39

Albumin, mg/dL 4.1±0.40 * 3.8±0.41 4.0±0.36 4.2±0.23 4.3±0.35

Total cholesterol, mg/dL 192±39 * 176±43 190±32 193±35 206±44

LDL-Cho, mg/dL 112±35 * 96±35 113±32 114±31 123±38

Triglycerides, mg/dL 119±77 127±86 113±71 122±86 116±63

HDL-Cho, mg/dL 56±17 * 52±20 55±15 55±16 60±18

Phosphorus, mg/dL 3.0±1.0 2.9±1.2 2.9±1.2 2.9±0.66 3.1±0.84

Blood urea nitrogen, mg/dL 18.4±16.2 21.2±24.8 16.2±10.6 16.8±9.6 20.1±16.7

Creatinine, mg/dL 1.1±1.8 1.3±2.1 0.85±0.68 1.02±1.78 1.31±2.27

Aspartate transaminase, IU/L 36.5±28.3 42.2±35.1 35.6±19.9 30.8±12.4 38.5±39.6

Alanine aminotransferase, IU/L 26.1±26.8 24.1±18.9 28.6±38.0 24.6±15.4 26.8±27.6

Blood glucose, mg/dL 142±51 154±75 142±41 136±45 135±38

hs-CRP, mg/dL 0.36±0.83 0.69±1.41 0.25±0.43 0.21±0.41 0.33±0.71

Hemoglobin, g/dL 13.8±1.9 * 13.0±2.0 13.9±1.9 14.1±1.7 14.2±2.0

Platelet count, ×104/μL 22.0±7.0 20.9±7.9 22.0±6.7 23.1±7.1 22.0±6.1

aPTT, s 28.3±5.0 30.2±6.4 27.8±4.6 27.7±4.4 27.8±4.1

PT-INR 1.04±0.35 1.1±0.41 1.0±0.32 1.0±0.36 1.0±0.33

eGFR, mL/min 76.4±27.6 77.5±34.4 77.8±36 76.4±22.8 73.7±28.3

Length of hospital stay (days) 26 (17-38) 30 (19-39) 26 (17-39) 25 (19-37) 25 (14-39)

5

at WORLD HLTH ORGANIZATION on June 23, 2013http://stroke.ahajournals.org/Downloaded from

Inoue et al: Serum Calcium, hematoma volume and severity in ICH, p6

Data are presented as means±SD for continuous variables, medians with interquartile ranges in

parentheses for discontinuous variables, and numbers of patients with percentages in parentheses for

the others. * p<0.05 among quartiles.

Data on total cholesterol and aPTT are missing in 1 patient, those on triglycerides in 4, LDL-Cho in 6,

HDL-Cho in 20, and phosphorus in 28.

Abbreviation: LDL-Cho, low-density-lipoprotein cholesterol; HDL-Cho, high-density-lipoprotein

cholesterol; hs-CRP, high sensitive C-reactive protein; aPTT, activated partial thromboplastin time;

PT-INR, prothrombin time-international normalized ratio; e-GFR, estimated glomerular filtration rate

6

at WORLD HLTH ORGANIZATION on June 23, 2013http://stroke.ahajournals.org/Downloaded from

Inoue et al: Serum Calcium, hematoma volume and severity in ICH, p7

7

Table S3. Multivariate analysis for outcomes at 30 days

Crude Adjusted: model 1 Adjusted: model 2

Patients

(%) OR (95% CI) p value OR (95% CI) p value OR (95% CI) p value

Modified Rankin Scale score 5-6

Q1 25 (43.9) 2.70 (1.22-6.21) 0.016 2.59 (0.99-7.14) 0.057 2.05 (0.59-7.67) 0.270

Q2 16 (21.3) 0.94 (0.41-2.18) 0.881 1.06 (0.39-2.94) 0.913 0.79 (0.22-2.98) 0.726

Q3 14 (21.2) 0.93 (0.40-2.21) 0.872 1.37 (0.49-3.96) 0.548 1.43 (0.40-5.43) 0.591

Q4 13 (22.4) 1 (Reference) … 1 (Reference) … 1 (Reference) …

Death

Q1 10 (17.5) 3.90 (1.12-18.2) 0.048 3.60 (0.34-48.3) 0.302 3.21 (0.12-112.0) 0.481

Q2 3 (4.0) 0.76 (0.14-4.26) 0.747 0.75 (0.08-6.97) 0.800 0.48 (0.02-12.8) 0.640

Q3 2 (3.0) 0.57 (0.07-3.57) 0.550 1.93 (0.15-24.2) 0.600 4.90 (0.25-170.6) 0.317

Q4 3 (5.2) 1 (Reference) … 1 (Reference) … 1 (Reference) …

Patients with prestroke mRS score ≥2 were excluded from the analysis.

Model 1: Adjusted by age, sex, prestroke statin use, intraventricular bleeding, aspartate transaminase,

and blood glucose in “Modified Rankin Scale score 5-6”.

Adjusted by age, sex, liver dysfunction, prestroke antiplatelet use, systolic blood pressure, albumin,

hemoglobin, and platelet count in “Death”.

Model 2: Adjusted by variables in Model 1 plus NIHSS score on admission and hematoma volume.

at WORLD HLTH ORGANIZATION on June 23, 2013http://stroke.ahajournals.org/Downloaded from