Association of Circulating Inflammatory Markers With...

Association of Circulating Inflammatory Markers WithRecurrent Vascular Events After Stroke

A Prospective Cohort Study

William Whiteley, MRCP; Caroline Jackson, PhD; Steff Lewis, PhD; Gordon Lowe, DSc;Ann Rumley, PhD; Peter Sandercock, FRCP; Joanna Wardlaw, FRCP;

Martin Dennis, FRCP; Cathie Sudlow, FRCP

Background and Purpose—Inflammatory markers may be associated with recurrent vascular events after stroke. Weaimed to determine the association between IL-6, C-reactive protein, fibrinogen and white cell count, with recurrentvascular events after stroke, and to compare the association between circulating inflammatory markers with the risk ofdeath from vascular vs nonvascular causes.

Methods—We prospectively recruited patients with acute stroke (n�817) and followed them for up to 4 years for the occurrenceof fatal or nonfatal recurrent stroke, myocardial infarction or fatal vascular events, and death from any cause (n�159).

Results—The delay to assessment was a median of 10 days. The adjusted incidence of the outcome cluster recurrent stroke,myocardial infarction or vascular death after stroke was significantly higher with higher levels of IL-6 (75th to 25th

percentile hazard ratio, 1.56; 95% CI, 1.37–1.77), C-reactive protein (75th to 25th percentile hazard ratio, 1.08; 95% CI,1.04–1.11), and fibrinogen (75th to 25th percentile hazard ratio, 1.45; 95% CI, 1.24–1.72). The associations betweeninflammatory markers and death were stronger than with recurrent vascular events. The associations of inflammatorymarkers with vascular and nonvascular deaths were similar.

Conclusions—Although inflammatory markers were associated with an increased risk of recurrent vascular events andvascular death after stroke, they were also associated with nonvascular causes of death, suggesting that inflammatorymarkers do not play a causal role specifically in the generation of recurrent vascular events after stroke. Future studiesof the prediction of recurrent vascular events after stroke should concentrate on clinical variables or different bloodmarkers. (Stroke. 2011;42:10-16.)

Key Words: inflammation � prognosis � stroke

In prospective studies of patients with previous stroke orTIA, increased levels of markers of acute inflammation,

C-reactive protein (CRP),1 IL-6,2 fibrinogen,1,3 and white cellcount,4 were associated with increased incidence of recurrentstroke and MI and the risk of death or disability at 6 monthsafter stroke.5 If high levels of acute-phase markers in the earlystages of stroke contribute to recurrent vascular events, thenwe hypothesized the association between inflammatory mark-ers with fatal recurrent vascular events would be strongerthan with other nonvascular deaths.

In a prospective cohort of patients with recent stroke, weaimed to estimate the association between levels of circulat-ing inflammatory markers and the incidence of “recurrentvascular events” (recurrent stroke, MI, and vascular death),and to compare the strength of the association between

inflammatory markers and the risk of death from vascular andnonvascular causes.

Materials and MethodsThe Edinburgh Stroke Study was a prospective, hospital-basedcohort study of stroke patients followed-up for recurrent stroke, MI,and death. We have described the methods and process of datacollection elsewhere.6 In brief, we recruited consenting patients withischemic stroke or intracerebral hemorrhage presenting to the West-ern General Hospital in Edinburgh between April 2002 and May2005. We assigned ischemic stroke subtypes according to the siteand size of the causative infarct, modified if necessary by imagingfindings with the Oxford community stroke project classification,and with a modified Trial of ORG10172 in Acute Stroke Treatmentclassification as described previously.6 We made a clinical assess-ment at baseline and contemporaneously drew blood for markers ofinflammation (CRP, IL-6, fibrinogen and white cell count, and

Received April 28, 2010; accepted July 29, 2010.From the Division of Clinical Neurosciences (W.W., C.J., S.L., P.S., J.W., M.D., C.S.), Western General Hospital, SFC Brain Imaging Research Centre

(J.W.), SINAPSE Collaboration, Institute of Genetics and Molecular Medicine (C.S.), University of Edinburgh, Edinburgh, UK; Division ofCardiovascular and Medical Sciences (G.L., A.R.), Royal Infirmary, University of Glasgow, Glasgow, UK.

The online-only Data Supplement is available at http://stroke.ahajournals.org/cgi/content/full/STROKEAHA.110.588954/DC1.Correspondence to William Whiteley, MRCP, Bramwell Dott Building, Department of Clinical Neurosciences, Western General Hospital, Edinburgh,

EH4 2XU, UK. E-mail [email protected]© 2010 American Heart Association, Inc.

Stroke is available at http://stroke.ahajournals.org DOI: 10.1161/STROKEAHA.110.588954

10

by guest on June 11, 2018http://stroke.ahajournals.org/

Dow

nloaded from


Dow

nloaded from


Dow

nloaded from


Dow

nloaded from


Dow

nloaded from


Dow

nloaded from


Dow

nloaded from


Dow

nloaded from


Dow

nloaded from


Dow

nloaded from


Dow

nloaded from


Dow

nloaded from


Dow

nloaded from


Dow

nloaded from


Dow

nloaded from


Dow

nloaded from


Dow

nloaded from


Dow

nloaded from


Dow

nloaded from


Dow

nloaded from


Dow

nloaded from

http://stroke.ahajournals.org/





















glucose). We obtained follow-up data for patients with multipleoverlapping methods. We defined recurrent vascular events as theoutcome cluster “recurrent fatal or nonfatal stroke, subsequent fatalor nonfatal MI, or other vascular death.” We defined other vasculardeath as deaths attributable to atherothromboembolic vascular dis-eases other than stroke or MI. For this study, we classified deathsattributable to the qualifying, although not recurrent, stroke (whichwere most often attributable to pneumonia) or gastrointestinalhemorrhage as nonvascular. We did not routinely record the occur-rence of infections or other complications between stroke onset andthe measurement of vascular outcome or death. The Lothian Re-search Ethics Committee approved the project.

Measurement of Blood MarkersA clinical laboratory measured total white cell count (BeckmanCoulter LH750 analyzer) and blood glucose (Vitros Chemistryanalyzer). Blind to clinical details, we measured CRP and fibrinogenin plasma by immunonephelometry (Prospec, Dade Behring MiltonKeynes) using the manufacturer’s reagents and standards. Weassayed IL-6 by enzyme-linked immunosorbent assay (R & DSystems). Intra-assay and interassay coefficients of variation were4.7% and 8.3%, 2.6% and 5.3%, and 7.5% and 8.9%, respectively.

Statistical AnalysisWe used Stata version 10 (Stata 2007) for analysis and prepared thearticle with reference to the Strengthening the Reporting of Obser-vational Studies in Epidemiology (STROBE)7 guidelines. We mea-sured survival to first recurrent vascular event and separately to eachof recurrent stroke, MI, and vascular death, censoring patients at theend of follow-up or nonvascular death. We compared the baselinecharacteristics of patients who experienced a recurrent vascular eventwith those who did not with univariable Cox regression analysis. Weexamined the relationships between inflammatory markers withcorrelation coefficients and used linear regression (after loge trans-formation of markers) to examine the relationship of markers withdelay to blood draw. We used Kaplan-Meier survival curves tocompare event-free survival between groups of patients defined bythirds of inflammatory biomarkers and compared curves with log-rank trend tests. We used Cox regression analysis to calculateunadjusted hazard ratios (HR) and 95% CI per unit increase inmarker levels. We built a multivariate Cox regression model toadjust for confounders, adding variables sequentially that wereassociated with recurrent vascular events in univariable analysis andhad data completeness �95%, keeping those variables that signifi-cantly improved the fit of the model (likelihood ratio test P�0.05).When the model was complete, we tested the proportional hazardsassumption and its goodness of fit. We looked for first-orderinteractions of inflammatory biomarker levels with other variables inthe final model by adding multiplicative terms.

We made further assessment of the marker most strongly associ-ated with recurrent vascular events (IL-6). We assessed the change indiscrimination after the addition of IL-6 to a model containing onlyclinical variables by calculating Harrell c-statistic for models withand without biomarkers. The c-statistic is analogous to the area undera receiver-operator curve for Cox regression models; a value of 0.5indicates no better discrimination than chance and a value of 1indicates perfect discrimination.

We replicated the analysis measuring time to death only, censor-ing at the end of the study. To adjust models examining the risk ofdeath, we used previously validated covariates.8 We also plotted, bythirds of marker levels, the competing risks of vascular deaths, deathattributable to the initial stroke, and death attributable to other causesusing the “stcompet” command,9 which calculates the cumulativeincidence of each outcome.

ResultsBaseline CharacteristicsEight-hundred seventy-seven of 1408 patients in the Edin-burgh Stroke Study (62%) gave consent and were available

for blood to be drawn for markers of inflammation. Of these,817 (93%) had a definite ischemic stroke, 17 (2%) had aprobable ischemic stroke, and 43 (5%) had a hemorrhagicstroke. Of those patients who had blood drawn for bloodmarkers, none was lost to follow-up for the outcomes ofdeath, recurrent stroke, or MI. We first assessed patientsclinically at a median of 10 days (interquartile range [IQR],3–21 days) after stroke onset and drew blood at a median of0 days (IQR, 0–3 days) after assessment. The delay toassessment was longer for outpatients seen in a clinic (me-dian, 19 days) than in those admitted to the stroke unit(median, 2 days). Patients who did not have blood drawnwere similar to those with blood drawn for biomarker data inage, gender distribution, and the proportions with hyperten-sion, peripheral or cardiac vascular disease, diabetes, or atrialfibrillation. On average, compared to those without bloodsampling, patients with blood samples had milder strokes(proportion of total anterior circulation stroke, 6.8% vs14.7%; P�0.001) than patients admitted to hospital, andthose with more severe symptoms were less likely to berecruited because of practical barriers to obtaining andprocessing research blood samples and obtaining informedconsent or assent.

During the 1866 person-years of follow-up (mean, 2.12years), 106 patients had a first recurrent stroke (92 ischemic,5 hemorrhagic, and 9 of uncertain type) and 34 had MI. Therewere 184 deaths: 113 from vascular causes (63 strokes, 35from cardiac causes, and 15 deaths from bowel ischemia,vascular dementia, and presumed vascular renal failure) and64 from other causes (33 cancers, 13 chest infections, 6 fromchronic obstructive pulmonary disease, and the rest frompancreatitis, bowel perforation, hip fracture, and extrapulmo-nary sepsis).

At the time of the clinical assessment of the index stroke,the median IL-6 was 4.0 pg/L (IQR, 2.4–7.2 pg/L), medianCRP was 3.5 mg/L (IQR, 1.4–9.7 mg/L), median fibrinogenwas 4.5 g/L (IQR, 3.8–5.4 g/L), median white cell count was8�109/L (IQR, 6.6–9.7 �109/L), and median glucose was5.6 mmol/L (IQR, 5–6.8 mmol/L). The correlation coeffi-cients between IL-6 and other variables were: CRP, 0.59;fibrinogen, 0.48; glucose, 0.06; and white cell count, 0.25.The correlation coefficients between loge marker levels anddelay from symptom onset to blood draw were 0.18 for IL-6,�0.12 for CRP, �0.11 for fibrinogen, �0.13 for white cellcount, and �0.06 for glucose. The relationships of averagemarker levels with the delay from stroke to blood draw werenot statistically significant after adjusting for the level ofbaseline neurological impairment and age.

Circulating Inflammatory Markers and RecurrentStroke, MI, and Vascular DeathThere was a significant increase in recurrent vascular eventsfor patients who were older, had a history of atrial fibrillationor heart failure, or had previous peripheral vascular disease,coronary heart disease, or stroke (Table 1). The log hazard ofstroke, MI, or vascular death increased with each third of IL-6and CRP, although not by thirds of glucose, fibrinogen, orwhite cell count.

Whiteley et al Inflammatory Markers and Vascular Events 11


Dow

nloaded from


Table 1. Baseline Characteristics of Stroke Patients

TotalRecurrent Stroke,

MI or Vascular DeathNo Vascular

EventUnivariate HR

(95% CI)

N 877 159 718

Demographic

Age, y (mean, SD) 71.4 (12.0) 73.6 (10.7) 70.9 (12.2) 1.02 (1.00–1.04)*

Male, N (%) 463 (52.8) 80 (50.3) 383 (53.3) 0.9 (0.6–1.2)

Laboratory results Median (IQR) Median (IQR) Median (IQR)

IL-6 (pg/mL) 4.0 (2.4–7.2) 4.8 (2.8–9.1) 3.8 (2.3–6.6)

CRP (mg/L) 3.5 (1.4–9.7) 5.9 (1.9–15.5) 3.3 (1.2–8.8)

Fibrinogen (g/L) 4.5 (3.8–5.4) 5.7 (3.9–5.7) 4.4 (3.8–5.4)

White cell count (�109/L) 8 (6.6–9.7) 8.4 (6.7–9.7) 7.9 (6.6–9.7)

Glucose (mmol) 5.6 (5–6.8) 5.7 (4.9–7.2) 5.6 (5–6.7)

Cholesterol (mmol/L) 5.1 (4.4–6.0) 4.9 (4.3–5) 5.1 (4.4–6)

Pathological type, index stroke N (%) N (%) N (%)

Definite ischemic 817 (93.2) 149 (93.7) 668 (93.0) 1.1 (0.6–2.1)†

Definite hemorrhagic 43 (4.9) 9 (5.7) 34 (4.7) 1.2 (0.6–2.4)†

Pathological type unknown 17 (1.9) 1 (0.6) 16 (2.2) 0.3 (0.1–1.9)†

Clinical stroke syndrome of index stroke (OCSP)

TACI 57 (6.8) 10 (6.3) 53 (7.4) 1.2 (0.6–2.2)‡

PACI 376 (45.1) 81 (50.9) 308 (42.9) 1.1 (0.8–1.6)‡

LACI 228 (27.3) 38 (23.9) 200 (27.9) 0.8 (0.5–1.1)‡

POCI 131 (15.7) 22 (13.8) 121 (16.9) 0.8 (0.5–1.3)‡

Uncertain subtype 42 (5.0) 8 (5.0) 36 (5.0) 0.9 (0.5–1.9)‡

Severity of index stroke

Cannot walk, cannot lift both arms 91 (10.4) 12 (7.6) 79 (11.0) 0.9 (0.5–1.5)

Cannot walk, can lift both arms 119 (13.6) 31 (19.5) 88 (12.3) 1.6 (1.1–2.4)

Can walk 664 (76.0) 115 (72.9) 549 (76.7) 0.7 (0.5–1.0)

Etiological classification of index stroke (TOAST)

Cardioembolic 117 (13.3) 28 (17.6) 89 (12.4) 1.2 (0.8–2.0)§

Large vessel disease 72 (8.2) 12 (7.6) 60 (8.4) 1.0 (0.7–1.6)§

Mixed aetiology 58 (6.6) 16 (7.6) 42 (5.9) 1.8 (1.2–2.7)§

Small vessel disease 178 (20.3) 24 (15.1) 154 (21.5) 0.8 (0.6–1.2)§

Unclassified after complete investigation 355 (40.5) 65 (40.9) 290 (40.4) 0.9 (0.7–1.3)§

Unclassified after incomplete investigation 40 (15.8) 14 (8.8) 83 (11.6) 0.8 (0.5–1.5)§

Risk factors

History of TIA 143 (16.3) 29 (18.2) 114 (15.9) 1.1 (0.8–1.7)

History of stroke 166 (18.9) 41 (25.8) 125 (17.4) 1.5 (1.1–2.2)

History of ischemic heart disease 242 (27.6) 62 (39.0) 180 (25.1) 1.9 (1.4–2.6)

History of peripheral vascular disease 69 (7.9) 22 (13.9) 47 (6.6) 2.0 (1.3–3.2)

Ipsilateral carotid stenosis �70 97 (12.5) 21 (14.9) 76 (12.0) 1.2 (0.8–2.0)

Ever had atrial fibrillation 168 (20.3) 42 (26.4) 126 (17.6) 1.7 (1.3–2.5)

Previous treated hypertension 467 (53.3) 96 (60.4) 371 (51.7) 1.4 (1.0–1.9)

Diabetes 110 (12.5) 26 (16.4) 84 (11.7) 1.4 (0.9–2.2)

Ever smoker 602 (69.8) 113 (71.1) 489 (69.5) 1.1 (0.8–1.5)

Heart failure 40 (4.58) 14 (8.7) 26 (3.6) 2.8 (1.6–4.8)

Any antiplatelet at baseline 369 (46.1) 11 (6.9) 33 (4.6) 1.7 (0.9–3.1)

Warfarin at baseline 43 (4.9) 32 (4.5) 11 (6.9) 1.5 (0.8–2.7)

Systolic BP (Mean, N of observations) 147.2 (874) 147.3 (159) 147.2 (715) 1.00 (0.99–1.01)*

Diastolic BP (Mean, N of observations) 80.0 (874) 80.1 (159) 80.0 (715) 1.00 (0.99–1.01)*

LACS indicates lacunar stroke; OCSP, Oxfordshire Community Stroke Project Classification (ischemic and probable); PACS, partialanterior circulation stroke; POCS, posterior circulation; TACS, total anterior circulation stroke; TOAST, Trial of ORG10172 in AcuteStroke Treatment.

*Per unit increase.†Vs other pathological types.‡Vs all others in OCSP classification.§Vs all others in TOAST classification.

12 Stroke January 2011


Dow

nloaded from


In unadjusted Kaplan-Meier survival analyses, patientssurvived free of recurrent vascular events for a shorter time inthe highest third of IL-6 (log-rank trend �2�13.2; P�0.0003;Figure 1) and CRP (log-rank trend �2�13.9; P�0.0002).This relationship did not reach statistical significance forfibrinogen (log-rank trend �2�2.8; P�0.09), glucose (log-rank trend �2�1.1; P�0.3), or white cell count (log-ranktrend �2�3.1; P�0.08). However, a linear model fit the datawell for each marker.

Table 2 shows the association between circulating inflam-matory markers and recurrent vascular events. In univariateanalyses, all markers except glucose were significantly asso-ciated with recurrent vascular events. The relative HR for anincrease of 1 pg/mL of IL-6 was 1.07 (95% CI, 1.04–1.10)per pg/mL. The unadjusted associations between IL-6 andrecurrent fatal or nonfatal stroke alone (HR, 1.04; 95% CI,1.00–1.08 per pg/mL) were weaker; however, the HR formg/L increase of CRP, g/L increase of fibrinogen, 1�109

increase in white cell count, or mmol/L increase of glucosewere not significantly different from 1. The unadjustedassociation between IL-6 and fatal or nonfatal MI alone (HR,1.09; 95% CI, 1.03–1.15) was stronger than for recurrentstroke alone.

We adjusted for the following confounders in the finalmodel: age, previous stroke or TIA or ischemic heart disease,current or previous atrial fibrillation, and cardiac failure.Adding markers of stroke severity (ie, ability to walk or liftarms off bed), blood pressure at assessment, stroke patholog-ical type, diabetes, carotid stenosis, delay to blood taking, orsmoking did not significantly improve models containing asingle inflammatory marker. After adjustment, there was stilla significant association between recurrent vascular eventsand increasing levels of IL-6, CRP, and fibrinogen (Table 2).In this cohort, those patients with highest blood levels (75th

percentile) of IL-6 had a 1.33-fold increase in the incidence ofrecurrent vascular events compared with those with thelowest levels (25th percentile). A similar relative increase inincidence was seen for fibrinogen (HR, 1.20), and lessincrease was seen for C-reactive protein (HR, 1.06).

We added markers sequentially as continuous variables, inorder of the strength of their association with recurrentvascular events, to a model containing only clinical variables(age, previous stroke or TIA or heart disease, current orprevious atrial fibrillation or cardiac failure). Addition of IL-6significantly improved the model (likelihood ratio test

0.75

1

ular

dea

th o

r MI

00.

5

recu

rren

t stro

ke, v

ascu

bottom third IL-6

middle third IL-6

log rank trend p=0.00030.25

0

Sur

viva

l fre

e of

top third IL-6

0

339561552492middle third IL-6038561862292bottom third IL-6

Number at risk

0 1 2 3 4Time (years)

116421702192top third IL-6

Figure 1. Unadjusted Kaplan-Meier sur-vival curve and life table for survival freefrom recurrent stroke, MI, or vasculardeath by one-third of IL-6.

Table 2. Association Between Marker Level and Recurrent Stroke, MI, or Vascular Death Assuming aLinear Association Between Marker Level and Log Hazards

Hazard Ratio Per Unit Increase* in Marker (95% CI) Hazard Ratio Comparing75th to 25st Percentile†

Adjusted‡Unadjusted Adjusted‡ Adjusted for All Markers§

IL-6 (pg/mL) 1.07 (1.04–1.10) 1.06 (1.03–1.09) 1.05 (1.01–1.09) 1.33 (1.15–1.53)

C-reactive protein (mg/L) 1.01 (1.00–1.01) 1.01 (1.00–1.01) 1.00 (1.00–1.03) 1.06 (1.02–1.09)

Fibrinogen (g/L) 1.16 (1.05–1.28) 1.12 (1.01–1.25) 1.02 (0.97–1.17) 1.20 (1.01–1.43)

White cell count (�109/L) 1.06 (1.02–1.10) 1.05 (1.00–1.11) 1.03 (0.97–1.09) 1.17 (0.98–1.38)

Glucose (mmol/L) 1.04 (0.99–1.09) 1.03 (0.98–1.08) 1.02 (0.97–1.08) 1.06 (0.97–1.15)

*For IL-6, pg/mL; CRP, mg/L; fibrinogen, g/L; white cell count, �109/L; glucose, mmol/L.†25th and 75th percentiles, respectively, for IL-6, 2.39 and 7.22 pg/mL; CRP, 1.39 and 9.65 mg/L; fibrinogen, 3.81 and 5.41 g/L;

white cell count, 6.6 and 9.7 �109/L; glucose, 5.0 and 6.8 mmol/L.‡Adjusted for confounders age, cardiac failure, atrial fibrillation (current or past), or previous stroke, TIA, peripheral vascular

disease, or MI.§Adjusted for all confounders in previous column and other markers.



Dow

nloaded from


�2�14.0; P�0.001), although further addition of CRP (like-lihood ratio test �2�0.3; P�0.56), fibrinogen (likelihoodratio test �2�0.2; P�0.68), white cell count (likelihood ratiotest �2�0.7; P�0.40), or glucose (likelihood ratio test�2�1.3; P�0.25) did not cause further improvement, proba-bly because the markers were correlated. After adjustment forall markers, only the association between IL-6 and recurrentvascular events remained statistically significant. The finalmodel including IL-6 fulfilled the proportional hazards as-sumption and fit the data well.

A model with only clinical variables (age, previous TIA,MI, or stroke, and atrial fibrillation) had a Harrell c-statisticof 0.62. When we added IL-6 to this model, the Harrellc-statistic increased by a small amount (to 0.64).

Modification of Associations by Baseline VariablesMultiplicative interaction terms between delay to bloodtaking after stroke, large vessel stroke vs all others (with amodified Trial of ORG10172 in Acute Stroke Treatmentalgorithm), age, ability to walk, and level of IL-6 did notmake important changes to the association between IL-6 andrecurrent vascular events (and none significantly improvedthe fit of the final Cox proportional hazards model). This

means that, for example for the variable delay to bloodtaking, we were unable to demonstrate a difference in theassociation between IL-6 and recurrent vascular events inthose where blood was taken early (�5 days) or later (�5days) after stroke. There were no significant 2-way interac-tions between other blood markers and IL-6.

Circulating Inflammatory Markers and DeathAll markers were significantly associated with an increasedrisk of death (Table 3). After adjustment for factors that areknown reliably to influence survival after the index stroke(age, being able to walk or talk, independence of dailyactivities before stroke, being able to lift arms from the bed),these associations remained statistically significant althoughattenuated. IL-6, CRP, fibrinogen, and glucose were morestrongly associated with death than with recurrent vascularevents, although white cell count was less strongly associ-ated. After additional adjustment for all other markers, onlythe associations of IL-6 and fibrinogen with death remainedstatistically significant. The association between higher levelsof IL-6, CRP, and fibrinogen and an increased incidence ofdeath was consistent for each of the separate causes of death(vascular deaths, deaths attributable to the initial stroke, and

.15

.2

ekorts xedni morf htaeD Bhtaed ralucsav rehto & IM ,ekorts tnerruceR A

0.0

5.1

1 2 3 4ncid

ence

.15

.2

1 2 3 4

C Non vascular deaths

bottom third IL-6

middle third IL-6

Cum

ulat

ive

in

0.0

5.1

0 1 2 3 4

top third IL-6

Time to death (years)

Figure 2. Unadjusted cumulative inci-dence curves of (A) death from recurrentstroke, MI, or other vascular causes, (B)death from the initial stroke with no evi-dence of recurrent stroke, or (C) othernonvascular deaths, by thirds of IL-6 esti-mated from a competing risk analysis.

Table 3. Association Between Marker Level and Death From Any Cause, Assuming a Linear AssociationBetween Marker Level and Log Hazards

Hazard Ratio Per Unit Increase* in Marker (95% CI) Hazard Ratio Comparing75th to 25th Percentile†

Adjusted‡Unadjusted Adjusted‡ Adjusted for All Markers§

IL-6 (pg/mL) 1.13 (1.10–1.15) 1.10 (1.07–1.12) 1.07 (1.04–1.12) 1.56 (1.37–1.77)

C-reactive protein (mg/L) 1.01 (1.00–1.01) 1.01 (1.00–1.01) 1.00 (1.00–1.81) 1.08 (1.04–1.11)

Fibrinogen (g/L) 1.37 (1.26–1.49) 1.26 (1.14–1.40) 1.14 (1.01–1.28) 1.45 (1.24–1.72)

White cell count (�109/L) 1.07 (1.02–1.12) 1.05 (1.00–1.11) 1.03 (0.97–1.09) 1.17 (1.00–1.37)

Glucose (mmol/L) 1.95 (1.02–1.10) 1.06 (1.02–1.11) 1.04 (0.99–1.09) 1.12 (1.03–1.21)

*For IL-6, pg/mL; CRP, mg/L; fibrinogen, g/L; white cell count. �109/L; glucose, mmol/L.†25th and 75th percentiles, respectively, for IL-6, 2.39 and 7.22 pg/mL; CRP, 1.39 and 9.65 mg/L; fibrinogen, 3.81 and 5.41 g/L;

white cell count 6.6 and 9.7 �109/L; glucose, 5.0 and 6.8 mmol/L.‡Adjusted for confounders age, ability to walk, living alone, independent before stroke, orientated to place, time, and person, and

able to lift arms from bed.§Adjusted for all confounders in previous column and other markers.



Dow

nloaded from


deaths attributable to other causes; Figure 2, data for CRP andfibrinogen not shown). When the cause of death was thequalifying stroke, patients in the top one-third of the IL-6distribution had the shortest survival time.

DiscussionIn this cohort consisting of stroke patients on the stroke unitassessed soon after onset and outpatients with milder strokesseen after a short interval, higher levels of IL-6, CRP, andfibrinogen were associated with a higher incidence of recur-rent stroke, MI, or vascular death, independent of atrialfibrillation, previous vascular events, and age. In addition,higher levels of each inflammatory marker were associatedwith a higher incidence of death from all causes, an associ-ation that was stronger than for all vascular events. Theassociation with recurrent stroke alone was weak for IL-6 anddid not reach statistical significance in this cohort for theother markers.

We found no consistent evidence of different strengths ofassociation between higher levels of IL-6 with large-vessel-type strokes vs other stroke subtypes, strokes of differentseverity, or different times from stroke onset to blood draw.Stroke patients had qualitatively similar associations betweenIL-6, CRP, and fibrinogen and deaths from vascular and fromnonvascular causes. However, there was a suggestion thatearly deaths from the index stroke might be more stronglyassociated with higher levels of inflammation.

Strengths and LimitationsThis study had a number of methodological strengths. Thecohort included both mild and severe strokes, we used severaloverlapping methods to ensure we detected all recurrentvascular events, we determined vital status at the end of thefollow-up period for all of the cohort, and data for allsuspected outcome events were checked by the study clini-cians, either directly or by review of the medical and imagingrecords. The majority of patients with recurrent strokesunderwent repeat brain imaging (93%).

We were unable to draw blood for inflammatory markersfrom all patients. The most common reasons for not drawingblood were either the patient did not consent or the practicalconstraints of inpatients (chiefly the working hours of re-search laboratories handling the samples). Patients withoutblood samples tended to have more severe strokes, butotherwise they were similar (there was no evidence of aninteraction between stroke severity and the association be-tween inflammatory markers and either recurrent vascularevents or death). These selection biases may have influencedthe results.

We drew blood as soon as possible after assessment(median of 2 days among patient admitted to hospital); hence,levels of IL-6 and CRP were higher than in previous studies(delay to blood draw in these studies was between 12 hoursand 30 days), possibly because of, in some cases, strokecomplications such as pneumonia or deep vein thrombosis.Because we were unable to adjust for these in our analysis,the observed association may have been attributable toconfounding by these complications. However, in the 60% ofpatients seen as outpatients who had milder strokes and

probably fewer infections or other complications, the mediantime to blood draw was 19 days. Despite this, we were unableto demonstrate effect modification by the time to blood drawafter stroke on the association between either IL-6 or CRPand recurrent vascular events, although statistical tests foreffect modification are of low power. It is also possible thatamong patients in whom the initial assessment was delayed,we may have overlooked some early recurrent strokes.

We did not find that large vessel stroke subtypes had astronger association between marker levels and recurrentevents. However, a relatively large number of strokes wereunclassified by the Trial of ORG10172 in Acute StrokeTreatment classification, so we cannot exclude the possibilitythat an association exists.

We used a competing risks survival analysis to examine theassociation of IL-6, fibrinogen, and CRP with the 3 maincauses of death: vascular, nonvascular, and deaths attribut-able to the initial stroke. It is possible that there was somemisclassification of the cause of death, particularly for deathsoccurring soon after stroke when accurate attribution of thecause of death is difficult, even if autopsy is performed.

The epidemiological association between inflammatorymarkers and recurrent vascular events appears consistent andstrong, and similar for IL-6 and CRP, but it was somewhatweaker for fibrinogen. However, the degree of extra clinicalutility obtained by adding an inflammatory marker to aclinical predictive model is not merely determined by the factthat it is a statistically independent predictor in multivariatemodels. The small increase in the c-statistic achieved byadding IL-6 to a model based on clinical variables makes itunlikely that it will add clinically useful prediction to modelsbased on variables that do not require blood draw. We did notcalculate a threshold of IL-6 that best-predicted recurrentvascular events because such calculations usually lead to abiased assessment of the strength of predictive ability ofmarkers and are unlikely to be replicated in validationcohorts.10

InterpretationThese data suggest that CRP or IL-6 may not have a majorcausal role in recurrent vascular events after stroke. It is morelikely that the observed association reflects an inflammatoryresponse either to atherosclerosis or to its risk factors, or to anas yet unidentified trigger. In support of this, studies that haveexamined functional CRP and IL-6 polymorphisms (whichproduce differences in baseline CRP or IL-6 levels) founddifferent polymorphisms were not associated with an in-creased risk of stroke11 or other occlusive vascular events.12

However, to determine reliably any causal relationship be-tween the physiological inflammatory response and recurrentvascular events in humans would require a randomized trialof an agent that was an effective anti-inflammatory but had nodirect effect on other vascular risk factors such as cholesterolor blood pressure.

GeneralizabilityOur finding that IL-6 showed the strongest associations withrecurrent vascular events and with death in this cohort ofstroke patients is consistent with recent reports from



Dow

nloaded from


population-based prospective studies.13,14 Our estimates ofthe association between CRP, IL-6, fibrinogen, and recurrentstroke1,15,16 and CRP, fibrinogen, and death16,17 are consistentwith previous studies (Supplementary Table I available on-line at http://stroke.ahajournals.org).

ConclusionsWe have demonstrated an association between higher levelsof IL-6, CRP, and fibrinogen and an increased incidence ofocclusive vascular events in patients after stroke. The asso-ciation between IL-6, CRP, and fibrinogen and fatal vascularand nonvascular events after stroke seems similar. Futurestudies of prediction of recurrent vascular events after strokemight better-study easily measured clinical variables or ex-amine the effect of different blood markers.

AcknowledgmentsThe authors thank Aidan Hutchison for his help with querying theEdinburgh Stroke Study (ESS) database, Mike McDowall and theEdinburgh staff of the Scottish Stroke Care Audit for storing andmanaging data for the ESS, the Wellcome Trust Clinical ResearchFacility in Edinburgh, the many stroke and research fellows whohelped to recruit patients, and particularly the patients themselves.

Sources of FundingDr Whiteley was supported by a Chief Scientist’s Office ClinicalAcademic Training Fellowship from the Scottish Government. DrSudlow was supported by a Clinician Scientist Award from theWellcome Trust (063668/Z/01/A) and is now funded by the ScottishFunding Council. Dr Jackson was supported by the Wellcome Trust(063668/Z/01/A) and a Binks Trust Research Fellowship. ProfessorWardlaw was supported by the Scottish Funding Council through theScottish Imaging Network, a Platform for Scientific ExcellenceCollaboration (www.sinapse.ac.uk). The imaging was conducted inthe SFC Brain Imaging Research Centre at the University ofEdinburgh (www.sbirc.ed.ac.uk), a SINAPSE centre.

DisclosuresNone.

References1. Woodward M, Lowe GDO, Campbell DJ, Colman S, Rumley A,

Chalmers J, Neal BC, Patel A, Jenkins AJ, Kemp BE, MacMahon SW.Associations of inflammatory and hemostatic variables with the risk ofrecurrent stroke. Stroke. 2005;36:2143–2147.

2. Welsh P, Lowe GDO, Chalmers J, Campbell DJ, Rumley A, Neal BC,MacMahon SW, Woodward M. Associations of proinflammatory cyto-kines with the risk of recurrent stroke. Stroke. 2008;39:2226–2230.

3. Rothwell PM, Howard SC, Power DA, Gutnikov SA, Algra A, van GJ,Clark TG, Murphy MF, Warlow CP. Fibrinogen concentration and risk ofischemic stroke and acute coronary events in 5113 patients with transientischemic attack and minor ischemic stroke. Stroke. 2004;35:2300–2305.

4. Grau AJ, Boddy AW, Dukovic DA, Buggle F, Lichy C, Brandt T, HackeW, for the CAPRIE Investigators. Leukocyte count as an independentpredictor of recurrent ischemic events. Stroke. 2004;35:1147–1152.

5. Whiteley W, Jackson C, Lewis S, Lowe G, Rumley A, Sandercock P,Wardlaw J, Dennis M, Sudlow C. Inflammatory markers and pooroutcome after stroke: a prospective cohort study and systematic review ofinterleukin-6. PLoS Med. 2009;6:e1000145.

6. Jackson CA, Hutchison A, Dennis MS, Wardlaw JM, Lewis SC, SudlowCLM. Differences between ischemic stroke subtypes in vascularoutcomes support a distinct lacunar ischemic stroke arteriopathy: a pro-spective, hospital-based study. Stroke. 2009;40:3679–3684.

7. von Elm E, Altman DG, Egger M, Pocock SJ, Gotzsche PC, Vanden-broucke JP. The strengthening the reporting of observational studies inepidemiology (STROBE) statement: guidelines for reporting observa-tional studies. Lancet. 2007;370:1453–1457.

8. Counsell C, Dennis M, McDowall M. Predicting functional outcome inacute stroke: comparison of a simple six variable model with otherpredictive systems and informal clinical prediction. J Neurol NeurosurgPsychiatry. 2004;75:401–405.

9. Coviello V. Cumulative incidence estimation in the presence of com-peting risks. Stata J. 2004;4:103–112.

10. Royston P, Moons KGM, Altman DG, Vergouwe Y. Prognosis andprognostic research: developing a prognostic model. BMJ. 2009;338:b604.

11. Ladenvall C, Jood K, Blomstrand C, Nilsson S, Jern C, Ladenvall P.Serum C-reactive protein concentration and genotype in relation to is-chemic stroke subtype. Stroke. 2006;37:2018–2023.

12. Zacho J, Tybjaerg-Hansen A, Jensen JS, Grande P, Sillesen H, Nor-destgaard BG. Genetically elevated C-reactive protein and ischemicvascular disease. N Engl J Med. 2008;359:1897–1908.

13. Danesh J, Kaptoge S, Mann AG, Sarwar N, Wood A, Angleman SB,Wensley F, Higgins JPT, Lennon L, Eiriksdottir G, Rumley A, WhincupPH, Lowe GDO, Gudnason V. Long-term interleukin-6 levels and sub-sequent risk of coronary heart disease: two new prospective studies and asystematic review. PLoS Med. 2008;5:e78.

14. Patterson CC, Smith AE, Yarnell JWG, Rumley A, Ben Shlomo Y, LoweGDO. The associations of interleukin-6 (IL-6) and downstream inflam-matory markers with risk of cardiovascular disease: The CaerphillyStudy. Atherosclerosis. 2010;209:551–557.

15. Campbell DJ, Woodward M, Chalmers JP, Colman SA, Jenkins AJ,Kemp BE, Neal BC, Patel A, MacMahon SW. Soluble vascular celladhesion molecule 1 and N-terminal pro-B-type natriuretic peptide inpredicting ischemic stroke in patients with cerebrovascular disease. ArchNeurol. 2006;63:60–65.

16. Elkind MS, Tai W, Coates K, Paik MC, Sacco RL. High-sensitivityC-reactive protein, lipoprotein-associated phospholipase A2, andoutcome after ischemic stroke. Arch Intern Med. 2006;166:2073–2080.

17. Di Napoli M, Papa F, Bocola V. C-reactive protein in ischemic stroke: anindependent prognostic factor. Stroke. 2001;32:917–924.



Dow

nloaded from


Sandercock, Joanna Wardlaw, Martin Dennis and Cathie SudlowWilliam Whiteley, Caroline Jackson, Steff Lewis, Gordon Lowe, Ann Rumley, Peter

Stroke: A Prospective Cohort StudyAssociation of Circulating Inflammatory Markers With Recurrent Vascular Events After

Print ISSN: 0039-2499. Online ISSN: 1524-4628 Copyright © 2010 American Heart Association, Inc. All rights reserved.

is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231Stroke doi: 10.1161/STROKEAHA.110.588954

2011;42:10-16; originally published online December 2, 2010;Stroke.

http://stroke.ahajournals.org/content/42/1/10World 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/2012/02/28/STROKEAHA.110.588954.DC3 http://stroke.ahajournals.org/content/suppl/2012/02/26/STROKEAHA.110.588954.DC2 http://stroke.ahajournals.org/content/suppl/2010/12/02/STROKEAHA.110.588954.DC1

Data 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:


Dow

nloaded from

http://stroke.ahajournals.org/content/42/1/10

http://stroke.ahajournals.org/content/suppl/2010/12/02/STROKEAHA.110.588954.DC1



http://www.ahajournals.org/site/rights/

http://www.lww.com/reprints

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


Supplementary table 1 .Thresholds from other studies applied to the Edinburgh Stroke Study (ESS), using the same analytical technique.

Study Marker Thresholds Outcome Reported association Association when threshold applied to ESS

Woodward 20051 CRP mg/L 1.14, 3.34 Recurrent stroke OR: 1.26 (0.98 to 1.61)* 1.18 (0.69 to 2.13)║ Fibrinogen g/L 3.32, 4.04 Recurrent stroke OR: 1.24 (0.97 to 1.59)* 0.97 (0.47 to 1.98) ║ Welsh 20082 IL-6 pg/ml 1.70, 2.94 Recurrent stroke OR: 1.31 (1.01 to 1.69)* 1.35 (0.61 to 2.97) ║ Di Napoli 200118 CRP mg/L 5, 33 Death or any vascular event HR: 2.89 (1.58 to 5.29)† 1.77 (1.41 to 2.21)** Fibrinogen g/L 3.78, 6.17 Death or any vascular event HR: 2.08 (1.19 to 3.62)† 1.85 (1.41 to 2.42) ** Rallidis 200819 CRP mg/l Per 1mg/l Death by hospital discharge OR: 1.20 (1.09 to 1.30) ‡ 1.01 (1.01 to 1.02) **†† Fibrinogen Per 0.1 g/L Death by hospital discharge OR: 1.18 (1.08 to 1.30) ‡ 1.02 (1.00 to 1.04) ** Elkind 200616 CRP mg/L 4.2, 10.3, 31.1 Recurrent vascular event HR: 1.86 (1.13 to 3.08)§ 1.91 (1.15 to 3.18) ║ CRP mg/L 4.2, 10.3, 31.1 Death HR: 4.50 (2.83 to 7.15)§ 3.21 (2.04 to 5.05) ** Campbell 200615 CRP mg/L 0.8, 1.8, 4.4. 92.1 Recurrent ischemic stroke OR: 1.00 (0.63 to 1.58) § 1.55 (0.14 to 3.30) ║ *Top versus bottom third, adjusted; †per third, unadjusted; ‡adjusted; § top to bottom quarter adjusted ║ adjusted for age, AF, prior TIA, stroke or MI and cardiac failure; ** adjusted for age, ability to walk, living alone, independent prior to stroke, orientated to place time and person, able to lift arms from bed†† death within first year only

(18) Di Napoli M, Papa F, Bocola V. Prognostic Influence of Increased C-Reactive Protein and Fibrinogen Levels in Ischemic Stroke. Stroke 2001 January 1;32(1):133-8.

(19) Rallidis LS, Vikelis M, Panagiotakos DB, Liakos GK, Krania E, Kremastinos DT. Usefulness of inflammatory and haemostatic markers to predict short-term risk for death in middle-aged ischaemic stroke patients. Acta Neurologica Scandinavica 2008;117:415-20.

7

Whiteley et al Inflammatory Markers and Vascular Events

Original Contributions

血炎性标记物与卒中后复发性血管事件的关系

一项前瞻性队列研究

Association of Circulating Inflammatory Markers With Recurrent Vascular Events After Stroke

A Prospective Cohort StudyWilliam Whiteley, MRCP; Caroline Jackson, Phd; Steff Lewis, PhD; Gordon LoweDSc; Ann Rumley, PhD;

Peter Sandercock, FRCP; Joanna Wardlaw, FRCP; Martin Dennis, FRCP; Cathie Sudlow, FRCP

背景和目的：炎性标记物可能与卒中后复发性血管事件相关。本研究旨在明确 IL-6、C 反应蛋白、纤维蛋白

原和白细胞计数与卒中后复发性血管事件的关系，并比较血炎性标记物与血管性和非血管性死亡风险的关联

强度。

方法：前瞻性纳入急性卒中患者 (n=817)，经四年随访，终点事件是致死性及非致死性复发性卒中、心肌梗塞、

致死性血管事件和其它原因导致的死亡 (n=159)。结果：评估延迟的中位数时间为 10 天。经校正，卒中后复发性卒中、心肌梗塞或致死性血管事件的发生率

高与较高的炎性标记物水平显著相关：IL-6 水平 ( 风险比 25%-75% 百分位数为 1.56，95% 可信区间 [CI] 为1.37-1.77)，C 反应蛋白水平 ( 风险比 25%-75% 百分位数为 1.08，95% CI 为 1.04-1.11)，纤维蛋白原水平 ( 风险比 25%-75% 百分位数为 1.45，95% CI 为 1.24-1.72)。与复发性血管事件相比，死亡与炎性标记物的相关性

更高。炎性标记物与血管性和非血管性死亡的相关性相似。

结论：炎性标记物除了与卒中后复发性血管事件和血管性死亡相关，也与非血管性死亡相关，由此提示炎性

标记物对于卒中后复发性血管事件的发生并非是特异性的因果关系。今后的研究应进一步探索临床特征或血

液中其它标记物能否预测卒中后复发性血管事件的发生。

关键词：炎症，预后，卒中

(Stroke. 2011;42:10-16.　复旦大学附属华山医院神经内科　董漪译程忻董强校 )

From the Division of Clinical Neurosciences (W.W., C.J., S.L., P.S., J.W., M.D., C.S.), Western General Hospital, SFC Brain Imaging Research Centre (J.W.), SINAPSE Collaboration, Institute of Genetics and Molecular Medicine (C.S.), University of Edinburgh, Edinburgh, UK; Division of Cardiovascular and Medical Sciences (G.L., A.R.), Royal Infirmary, University of Glasgow, Glasgow, UK.

The online-only Data Supplement is available at http://stroke.ahajournals.org/cgi/content/full/STROKEAHA.110.588954/DC1.Correspondence to William Whiteley, MRCP, Bramwell Dott Building, Department of Clinical Neurosciences, Western General Hospital, Edinburgh,

EH4 2XU, UK. E-mail [email protected]© 2010 American Heart Association, Inc.

既往的前瞻性研究认为，卒中或 TIA 患者的急

性炎性标记物水平增高，包括 C 反应蛋白 (CRP)[1]、

IL-6[2]、纤维蛋白原 [1,3] 和白细胞 [4]，并且其与卒中

后 6 月内的复发性卒中、心肌梗塞、死亡或致残相

关 [5]。如果卒中早期急性炎性标记物的高水平可能

促使复发性血管事件的发生，那么我们推测炎性标

记物与血管性死亡的关联性应强于其与非血管性死

亡的关联性。

我们的研究目的是在近期有卒中史的前瞻性队

列中，评估血炎性标记物与复发性血管事件 ( 复发

性卒中、心肌梗塞、血管性死亡 ) 发生之间的关系，

比较炎性标记物与血管性和非血管性死亡风险的相

关程度。

材料与方法爱丁堡卒中研究是一项前瞻性、以医院为基础

的的卒中病人队列研究，随访卒中后患者是否发生

复发性卒中、心肌梗塞或死亡。我们已发表了该研

究的方法及数据收集过程 [6]。简而言之，我们在获

取知情同意的前提下，纳入了 2002 年 4 月至 2005年 5 月在爱丁堡 Western 总医院就诊的缺血性和出

血性卒中患者。根据梗塞部位和病灶大小，进行缺

8

Stroke January 2011

血性卒中亚型分型，并在必要的情况下，根据牛津

社区卒中项目 (OCSP) 和之前所述的急性卒中治疗

ORG10172 临床研究 [6] 对分型进行修正。我们对基

线的情况进行了临床评估，并同时抽血测定炎性标

记物 (CRP，IL-6，纤维蛋白原，白细胞计数和血糖 )。我们通过多重方法进行患者随访，将复发性血管事

件定义为 “ 再发致死性或非致死性卒中，继发致死

性或非致死性心肌梗塞，或其它血管事件所致死亡 ”，其中其它血管事件所致死亡是指除心肌梗塞及卒中

以外的动脉粥样硬化血栓栓塞性疾病所致的死亡。

本研究的死因分类中，卒中后死亡 ( 大多是由于肺

炎 ) 或消化道出血所致的死亡均归为非血管性死亡。

本研究并未常规记录患者从卒中发病至发生血管性

事件或死亡期间，并发的感染或其他并发症的情况。

本项目经 Lothian 研究伦理委员会批准。

血生物学标记物测定

临床实验室测定全血白细胞计数 (Beckman Cou-l ter LH750分析仪 )和血糖 (Vitros Chemistry分析仪 )。按生产厂家的试剂及标准流程，采用免疫比浊法

(Prospec, Dade Behring Milton Keynes) 测定血浆 CRP及纤维蛋白原，采用酶联免疫吸附法 (R&D Systems)测定 IL-6。每次测定间和测定内部的变异系数分别

为 4.7% 和 8.3%，2.6% 和 5.3%，7.5% 和 8.9%。

统计学分析

我们采用第 10 版 Stata(Stata 2007) 软件，参照

流行病学观察性研究的强化报告指南 (STROBE)[7] 整

理分析数据。我们评估患者的生存情况至首次再发

血管性事件，分别评估其发生复发性卒中、心肌梗

塞及血管性死亡时的情况，跟踪患者至随访结束或

发生非血管性死亡。采用单因素 Cox 回归分析比较，

再发血管性事件的患者与未再发患者之间基线特征

的不同。分析炎性标记物与关联系数之间的关系，

并采用线性回归模型 ( 将炎性标记物数据 loge 转化

后 ) 分析炎性标记物与抽血延迟之间的关系。无事

件再发患者的炎性标记物数据按照三分法分组，采

用 Kaplan-Meier 生存曲线比较三组间的差异，并进

行组间的时序趋势检验。各组炎性标记物水平采用

Cox 回归分析，计算未经校正的危险比 (hazard ra-tios，HR) 及 95% 可信区间 (CI)。建立多因素 Cox 回

归分析模型，将单因素分析得出的与复发性血管事

件相关的变量依次纳入模型，校正各混淆因素，确

保数据完整性 >95%，使这些变量显著提高模型的拟

合度 ( 似然比检验 P<0.05)。模型建立后，检验其比

例风险假设和拟合优度。通过加入倍增项，探索炎

性生物标记水平与纳入最终模型的其它变量之间的

第一层关联。

我们进一步评估与复发性血管事件相关性最强

的炎性标记物 (IL-6)。将 IL-6 纳入仅含临床变量的

模型后，通过计算含或不含生物标记物的 Harrell c统计量，评价其差异。此 c 统计量等同于 Cox 回归

模型中 ROC 曲线下面积，若数值为 0.5 时提示为无

明显相关，若数值为 1，提示相关性强。

我们对发病至死亡的时间再次分析，监测至本

研究结束时，并使用既往验证的共变量校正死亡危

险因素的模型 [8]。我们还对标记物水平按照三分法

分组，应用竞争风险分析，采用“stcompet”方法 [9]，

计算得出的血管性死亡、首发卒中导致的死亡以及

其它原因导致死亡的累积发生率。

结果基线特征

爱丁堡卒中研究的 1408 例患者中，877 例 (62%)患者签署了知情同意书，并抽血检测炎性标记物。

其中，817 例患者 (93%) 确诊为缺血性卒中，17 例

(2%) 为可疑缺血性卒中，43 例 (5%) 为出血性卒中。

所有抽血的患者无一失访，随访的终点事件是死亡、

卒中再发和心肌梗塞。我们于卒中发病后中位数时

间 10 天对患者进行首次临床评估 ( 四分位数 [IQR] 3-21天 )以及中位数时间 0天进行抽血 (IQR，0-3天 )。门诊患者首次评估的延迟时间 ( 中位数，19 天 ) 较住院患者 ( 中位数，2 天 ) 更长。未抽血与抽血的患

者在年龄和性别分布、高血压、外周或心血管病变、

糖尿病及房颤患病比例方面相似。平均来说，较那

些未抽血的患者，抽血的患者较住院患者的卒中程

度轻 ( 前循环卒中比例 6.8% vs 14.7% ；P=0.001)，而症状较重的患者则由于临床流程而很难留取血样

本或签署知情同意，因此较少纳入本研究。

经 1866 人年的随访 ( 平均 2.12 年 )，106 例患

者出现了第一次再发卒中 (92 例缺血性卒中，5 例出

血性卒中及 9 例不明类型的卒中 )，34 例发生了心

肌梗塞。死亡病例 184 例：113 例死于血管性原因 (63例卒中，35 例心源性，15 例肠缺血、血管性痴呆、

推测的血管源性肾功能衰竭 )，64例死于其他原因 (33例癌症，13 例肺部感染，6 例慢性阻塞性肺病，其

余死于胰腺炎、肠穿孔、股骨骨折、肺外败血症 )。临床评估首次卒中时，IL-6 中位数水平 4.0 pg/L

9


表 1　卒中患者的基线特征再发卒中、心肌梗塞或单因素 HR 总计其它血管性死亡非血管性事件 (95% CI)N 877 159 718 人口学特点　年龄，岁 ( 平均值，标准差 ) 71.4 (12.0) 73.6 (10.7) 70.9 (12.2) 1.02 (1.00-1.04)*　男性，人数 (%) 463 (52.8) 80 (50.3) 383 (53.3) 0.9 (0.6-1.2)实验室指标中位数 ( 四分位数 ) 中位数 ( 四分位数 ) 中位数 ( 四分位数 ) 　IL-6 (pg/mL) 4.0 (2.4-7.2) 4.8 (2.8-9.1) 3.8 (2.3-6.6) 　CRP (mg/L) 3.5 (1.4-9.7) 5.9 (1.9-15.5) 3.3 (1.2-8.8) 　纤维蛋白原 (g/L) 4.5 (3.8-5.4) 5.7 (3.9-5.7) 4.4 (3.8-5.4) 　白细胞计数 (×109/L) 8 (6.6-9.7) 8.4 (6.7-9.7) 7.9 (6.6-9.7) 　血糖 (mmol) 5.6(5-6.8) 5.7(4.9-7.2) 5.6(5-6.7) 　胆固醇 (mmol/L) 5.1 (4.4-6.0) 4.9 (4.3-5) 5.1 (4.4-6) 病理类型，首次卒中 N (%) N (%) N (%) 　确诊缺血性卒中 817 (93.2) 149 (93.7) 668 (93.0) 1.1 (0.6-2.1)†　确诊出血性卒中 43 (4.9) 9 (5.7) 34 (4.7) 1.2 (0.6-2.4)†　病理类型不明 17 (1.9) 1 (0.6) 16 (2.2) 0.3 (0.1-1.9)†首次卒中临床卒中综合症 (OCSP) 　TACS 57 (6.8) 10 (6.3) 53 (7.4) 1.2 (0.6-2.2)‡　PACS 376 (45.1) 81 (50.9) 308 (42.9) 1.1 (0.8-1.6)‡　LACS 228 (27.3) 38 (23.9) 200 (27.9) 0.8 (0.5-1.1)‡　POCS 131 (15.7) 22 (13.8) 121 (16.9) 0.8 (0.5-1.3)‡　分类不明 42 (5.0) 8 (5.0) 36 (5.0) 0.9 (0.5-1.9)‡首次卒中严重程度　不能行走，不能上举双臂 91 (10.4) 12 (7.6) 79 (11.0) 0.9 (0.5-1.5)　不能行走，能上举双臂 119 (13.6) 31 (19.5) 88 (12.3) 1.6 (1.1-2.4)　可以行走 664 (76.0) 115 (72.9) 549 (76.7) 0.7 (0.5-1.0)首次卒中病因分类 (TOAST) 　心源性 117 (13.3) 28 (17.6) 89 (12.4) 1.2 (0.8-2.0)§　大血管性 72 (8.2) 12 (7.6) 60 (8.4) 1.0 (0.7-1.6)§　混合病因性 58 (6.6) 16 (7.6) 42 (5.9) 1.8 (1.2-2.7)§　小血管性 178 (20.3) 24 (15.1) 154 (21.5) 0.8 (0.6-1.2)§　完善检查仍无法分类 355 (40.5) 65 (40.9) 290 (40.4) 0.9 (0.7-1.3)§　未完成检查无法分类 40 (15.8) 14 (8.8) 83 (11.6) 0.8 (0.5-1.5)§危险因素　TIA 史 143 (16.3) 29 (18.2) 114 (15.9) 1.1 (0.8-1.7)　卒中史 166 (18.9) 41 (25.8) 125 (17.4) 1.5 (1.1-2.2)　缺血性心脏病史 242 (27.6) 62 (39.0) 180 (25.1) 1.9 (1.4-2.6)　周围血管病变史 69 (7.9) 22 (13.9) 47 (6.6) 2.0 (1.3-3.2)　同侧颈内动脉狭窄 >70% 97(12.5) 21(14.9) 76(12.0) 1.2(0.8-2.0)　曾发生过房颤 168 (20.3) 42 (26.4) 126 (17.6) 1.7 (1.3-2.5)　既往降压治疗 467 (53.3) 96 (60.4) 371 (51.7) 1.4 (1.0-1.9)　糖尿病史 110 (12.5) 26 (16.4) 84 (11.7) 1.4 (0.9-2.2)　曾吸烟 602 (69.8) 113 (71.1) 489 (69.5) 1.1 (0.8-1.5)　心力衰竭 40 (4.58) 14 (8.7) 26 (3.6) 2.8 (1.6-4.8)　基线抗血小板治疗 369 (46.1) 11 (6.9) 33 (4.6) 1.7 (0.9-3.1)　基线华法林治疗 43 (4.9) 32 (4.5) 11 (6.9) 1.5 (0.8-2.7)　收缩压 ( 平均值，测量次数 ) 147.2 (874) 147.3 (159) 147.2 (715) 1.00 (0.99-1.01)*　舒张压 ( 平均值，测量次数 ) 80.0 (874) 80.1 (159) 80.0 (715) 1.00 (0.99-1.01)*LACS，腔隙性梗塞；OCSP，牛津社区卒中项目分类 ( 缺血性和可能 ) ；PACS，部分前循环梗塞；POCS，后循环梗塞；TACS，完全前循环梗

塞；TOAST，急性卒中治疗 ORG10172 研究。

* 每单位增加。

† 与其它病理类型相比。

‡ 与其它 OCSP 分类相比。

§ 与其它 TOAST 分类相比。

10

Stroke January 2011

(IQR，2.4-7.2 pg/L)， CRP 中位数水平 3.5 mg/L(IQR，1.4-9.7 mg/L)，纤维蛋白原中位数水平 4.5 g/L(IQR，3.8-5.4 g/L)，白细胞计数中位数水平 8×109/L(IQR，6.6-9.7×109/L)，血糖中位数水平 5.6 mmol/L(IQR，5-6.8 mmol/L)。IL-6 与其它变量的相关性系数为：

与 CRP 0.59 ；与纤维蛋白原 0.48 ；与血糖 0.06 ；与

白细胞计数 0.25。上述指标的 loge 值与起病到抽

血之间的延误时间之间的相关性系数为 IL-6 0.18，CRP –0.12，纤维蛋白原 –0.11，白细胞计数 –0.13，血糖 –0.06。经年龄及基线神经系统损伤程度校正后，

上述指标的平均值与起病到抽血间的延误时间之间

的相关性均未达到统计学意义。

血炎性标记物与复发性卒中、心肌梗塞和血管性死亡

在老年、有房颤或心衰史、既往有外周血管疾病、

冠心病史、或卒中史的患者更易发生复发性血管事

件 ( 表 1)。随着 IL-6 及 CRP 水平 ( 按三分法 ) 的增

高，卒中、心肌梗塞或血管性死亡的风险 (log) 也增

加，但在按照三分法分组的血糖、纤维蛋白原及白

细胞计数上并没有得出如此结论。

未经校正的 Kaplan-Meier 生存分析发现，无复

发性血管事件时间最短的患者的 IL-6 水平 ( 时序趋

势检验 χ2=13.2; P=0.0003 ；图 1) 及 CRP 水平 ( 时序趋势检验 χ2=13.9 ；P=0.0002) 均位于最高的三分

之一组。但在纤维蛋白原 ( 时序趋势检验 χ2=2.8 ；

P=0.09)、血糖 ( 时序趋势检验 χ2=1.1 ；P=0.3)、及白

细胞计数 ( 时序趋势检验 χ2=3.1 ；P=0.08) 中这种关

联性未达到统计学意义，但各炎性标记物在线性模

型中拟合度好。

表 2 显示血炎性标记物与复发性血管事件的关

系。单因素分析发现，所有标记物除血糖外均与复

发性血管事件显著相关。IL-6 水平每增加 1 pg/mL，其相对 HR 是 1.07(95% CI，1.04-1.10)。IL-6 与复发

性致死性或非致死性卒中的未经校正的相关性不强

图 1　IL-6 水平三分法分组后无再发卒

中、心肌梗塞或血管性死亡的未经校

正的 Kaplan-Meier 生存曲线和生存表。

无再发卒中、血管性死亡或心肌梗塞

IL-6 水平最低三分之一组

IL-6 水平中间三分之一组

IL-6 水平最高三分之一组

时序趋势检验 p=0.0003

IL-6 水平最低三分之一组IL-6 水平中间三分之一组IL-6 水平最高三分之一组

时间 ( 年 )

表 2　假定标记物水平与风险比的 Log 值呈线性相关，标记物水平与复发性卒中、心肌梗塞或血管性死亡之间的相关性比较第 75 与第 25 标记物水平每单位增加的风险比 * (95% CI) 百分位数的风险比 † 未校正经校正 ‡ 所有标记物均经校正 § 经校正 ‡IL-6 (pg/mL) 1.07 (1.04 -1.10) 1.06 (1.03-1.09) 1.05 (1.01 -1.09) 1.33 (1.15-1.53)C 反应蛋白 (mg/L) 1.01 (1.00-1.01) 1.01 (1.00-1.01) 1.00 (1.00-1.03) 1.06 (1.02-1.09)纤维蛋白原 (g/L) 1.16 (1.05-1.28) 1.12 (1.01-1.25) 1.02 (0.97-1.17) 1.20 (1.01-1.43)白细胞计数 (×109/L) 1.06(1.02-1.10) 1.05 (1.00-1.11) 1.03 (0.97-1.09) 1.17 (0.98-1.38)血糖 (mmol/L) 1.04 (0.99-1.09) 1.03 (0.98-1.08) 1.02 (0.97-1.08) 1.06 (0.97-1.15)* 单位：IL-6，pg/mL ；CRP，mg/L ；纤维蛋白原，g/L ；白细胞计数，×109/L ；血糖，mmol/L。† 第 25 和第 75 百分位数分别为：IL-6，2.39 和 7.22 pg/mL；CRP，1.39 和 9.65 mg/L；纤维蛋白原，3.81 和 5.41 g/L；白细胞计数，6.6 和 9.7×109/L；

血糖，5.0 和 6.8 mmol/L。‡ 校正下列混淆因素：年龄、心衰、房颤 ( 现有或既往 )、既往卒中、TIA、周围血管疾病或心梗史。

§ 校正所有混淆因素，包括上述提及的及其它所有炎性标记物。

11


(HR，1.04 ；95% CI，1.00-1.08 每 pg/mL)。但 CRP水平每增加 1 mg/L、纤维蛋白原水平每增加 1 g/L、白细胞计数水平每增加 1×109 及血糖水平每增加 1 mmol/L，其 HR 与 1 无显著差异。IL-6 与致死性或

非致死性心肌梗塞的未经校正的相关性 (HR，1.09 ；

95% CI，1.03-1.15) 较复发性卒中的更密切。

在最终模型中，我们对下列混淆因素进行校正：

年龄、既往卒中或 TIA 或缺血性心脏病史、目前或

既往房颤史、心衰史。纳入卒中严重程度 ( 例如能

否行走或双手能否抬离床面 )、评估时血压水平、卒

中病理类型、糖尿病、颈动脉狭窄、延迟抽血时间

或吸烟等参数并不能显著优化含单一炎性标记物的

模型。经校正，复发性血管事件与 IL-6、CRP 和纤

维蛋白原水平的增高仍存在显著相关性 ( 表 2)。在

本研究队列中，IL-6 水平最高组 ( 第 75 百分位数 )患者的复发性血管事件的发生率较 IL-6 水平最低组

( 第 25 百分位数 ) 增加 1.33 倍。分析纤维蛋白原水

平 (HR 1.20) 也发现类似地发生率增加的现象，而在

C 反应蛋白的分析 (HR 1.06) 中，发生率增加较少。

我们将炎性标记物作为连续变量，依次纳入仅

包含临床变量 ( 年龄、既往卒中或 TIA 或心脏病史、

目前或既往房颤史、心衰史 ) 的模型，以增强其与

复发性血管事件关联的强度。纳入 IL-6 可显著改善

模型的拟合度 ( 似然比检验 χ2=14.0 ；P<0.001)，而

纳入 CRP( 似然比检验 χ2=0.3 ；P=0.56)、纤维蛋白

原 ( 似然比检验 χ2=0.2 ；P=0.68)、白细胞计数 ( 似然

比检验 χ2=0.7 ；P=0.40) 或血糖 ( 似然比检验 χ2=1.3 ；

P=0.25) 均无法进一步优化模型，可能是这些炎性标

记物互相相关。所有炎性标记物经校正，只有 IL-6与复发性血管事件显著相关，且有统计学意义。纳

入 IL-6 的最终模型符合比例风险假定，拟合度好。

仅含临床参数 ( 年龄、既往 TIA、心肌梗塞和

房颤史 ) 模型的 Harrell c 统计量为 0.62。当我们将

IL-6 纳入本模型中，Harrell c 统计量稍增加至 0.64。

表 3　假定标记物水平与风险比的 Log 值呈线性相关，标记物水平与任何原因死亡的相关性比较第 75 与第 25 标记物水平每单位增加的风险比 * (95% CI) 百分位数的风险比 † 未校正 ‡ 经校正 ‡ 所有标记物均经校正 § 经校正 ‡IL-6 (pg/mL) 1.13 (1.10 -1.15) 1.10 (1.07-1.12) 1.07 (1.04 -1.12) 1.56 (1.37-1.77)C 反应蛋白 (mg/L) 1.01 (1.00-1.01) 1.01 (1.00-1.01) 1.00 (1.00-1.81) 1.08 (1.04-1.11)纤维蛋白原 (g/L) 1.37 (1.26-1.49) 1.26 (1.14-1.40) 1.14 (1.01-1.28) 1.45 (1.24-1.72)白细胞计数 (×109/L) 1.07 (1.02-1.12) 1.05 (1.00-1.11) 1.03 (0.97-1.09) 1.17 (1.00-1.37)血糖 (mmol/L) 1.95 (1.02-1.10) 1.06 (1.02-1.11) 1.04 (0.99-1.09) 1.12 (1.03-1.21)* 单位：IL-6，pg/mL ；CRP，mg/L ；纤维蛋白原，g/L ；白细胞计数，×109/L ；血糖，mmol/L。† 第 25 和第 75 百分位数分别为：IL-6，2.39 和 7.22 pg/mL；CRP，1.39 和 9.65 mg/L；纤维蛋白原，3.81 和 5.41 g/L；白细胞计数，6.6 和 9.7×109/L；

血糖，5.0 和 6.8 mmol/L。‡ 校正下列混淆因素：年龄、可行走、生活自理、卒中前自理能力、时间、地点及人物定向力和双手能抬离床面。

§ 校正所有混淆因素，包括上述提及的及其它所有炎性标记物。

图 2　IL-6 水平三分法分组，竞争风险分析

显示未经校正的累积发生率曲线：(A) 因复

发性卒中、心肌梗塞或其它血管性原因导

致的死亡；(B) 首次卒中导致的死亡，而无

卒中再发的证据；(C) 其它非血管性死亡。

累积发生率

距离死亡的时间 ( 年 )

A 复发性卒中、心肌梗塞或其它血管性原因导致的死亡 B 首次卒中导致的死亡

C 非血管性死亡

IL-6 水平最低三分之一组

IL-6 水平中间三分之一组

IL-6 水平最高三分之一组

12

Stroke January 2011

基线变量相关性的修正

卒中后抽血延迟时间、大血管性卒中与其它类

型相比 ( 按改良 TOAST 分型 )、年龄、行走能力与

IL-6 水平的倍增交互作用，并不影响 IL-6 水平与复

发性血管事件的相关性 ( 无一显著优化最终 Cox 比

例风险模型的拟合度 )。这就意味着，例如卒中后

抽血延迟时间的不同，不论早 (<5 天 ) 或晚 (>5 天 )抽血都不能改变 IL-6 与复发性血管事件的相关性。

IL-6 与其它血标记物之间并无明显的相互影响。

血炎性标记物与死亡

所有炎性标记物与死亡风险的增加均显著相关

( 表 3)。校正首次卒中后公认的影响生存的因素 ( 年龄、行走或交谈的能力、卒中前生活独立能力、双

手能抬离床面 ) 后，减弱了这种相关性，但仍具有

统计学差异。其中 IL-6、CRP、纤维蛋白原和血糖

与死亡的相关性强于复发性血管事件，而白细胞计

数与死亡的相关性较弱。校正所有炎性标记物后，

仅 IL-6 和纤维蛋白原与死亡的相关性仍存在统计学

差异。IL-6、CRP、纤维蛋白原的高水平与死亡发生

率的增加相关，且对于各种死因 ( 血管性死亡、首

发卒中后的死亡以及其它原因的死亡；图 2 未显示

CRP 与纤维蛋白原的数据 )。若死因为卒中，IL-6在最高三分之一水平的患者，其生存期最短。

讨论本队列研究纳入的研究对象包括发病后迅速就

诊于卒中单元的卒中患者和发病后较短时间内就诊

于门诊的较轻的卒中患者，发现高水平 IL-6、CRP和纤维蛋白原与复发性卒中、心肌梗塞或血管性死

亡的发生率增高相关，而与房颤、既往血管性事件

和年龄无关。此外，各项炎性标记物高水平与各种

死亡的发生率增高相关，并且这种相关性高于与所

有血管性事件的相关性。IL-6 与复发性卒中的相关

性较弱，其它标记物与复发性卒中的相关性在本队

列研究中未达到统计学意义。

我们发现高水平 IL-6 与大血管性卒中相比其它

类型卒中、卒中的严重程度、卒中后不同的抽血时

间间无明确的不同程度的相关性。卒中患者的 IL-6、CRP、纤维蛋白原水平均与血管性死亡和非血管性

死亡的相关性较好，提示首次卒中后早期死亡很可

能与高水平的炎性反应相关。

研究优势和缺陷

本研究的方法学具有一定的优势，其队列研究

中包含了轻度和重度的卒中患者，我们通过数种交

叉的方法以确定我们随访获得了所有复发性血管事

件。我们对队列中所有的研究对象在随访结束时确

定其生存状态，所有可疑的终点事件的数据都会由

临床研究者直接或通过查阅病史或影像学资料来明

确。绝大部分复发性卒中的患者都复查了头颅影像

学检查 (93%)。我们无法从所有患者中抽血检测炎性标志物。

最常见的原因是由于未能获得患者的知情同意或住

院病人受到了相关工作流程的限制 ( 主要是实验室

工作时间处理样品的限制 )。未行抽血检查的卒中患

者中重症比例较高，而其它情况相似 ( 目前尚无证

据提示炎性标记物与复发性血管事件或死亡的相关

性与卒中的严重程度存在相互作用 )，但这种选择偏

倚可能会影响本研究结果。

我们通常在评估后尽快抽血 ( 住院患者的中位

数时间为 2 天 ) ；本研究所得出的 IL-6 及 CRP 水平

高于既往的研究 ( 既往类似研究抽血延迟时间在 12小时至 30 天内 )，可能由于一些病例存在卒中后肺

炎、深静脉血栓等并发症。由于我们无法在分析中

校正这些因素，观察到的相关性也可能归因于这些

并发症的混淆影响。然而，在 60% 的轻症卒中的门

诊病人中很少存在感染或其它并发症，中位数抽血

延迟时间是 19 天。尽管如此，我们无法显示卒中后

不同的抽血时间对于 IL-6 或 CRP 水平与复发性血管

事件的相关性的影响，虽然统计学检验显示这种影

响很小。当然也存在由于卒中诊断的延迟，可能漏

诊了部分早期复发的卒中。

我们未发现大血管卒中与炎性标记物水平及

复发性血管事件之间的联系更密切。然而，按照

TOAST 分型，相当部分的卒中归于不明原因的卒中，

故尚不能除外这种联系的可能性。

我们采用竞争风险生存分析的方法检测 IL-6、纤维蛋白原和 CRP 与三种主要死亡原因：血管性死

亡、非血管性死亡及死于初发的卒中之间的相关性。

可能存在某些死亡原因错误分类的情况，尤其是卒

中后迅速死亡的患者，即使行尸检也难以对死因进

行精确分类。

炎性标记物如 IL-6、CRP 与复发性血管事件的

流行病学相关性似乎比较密切和肯定，而纤维蛋白

原的这种相关性较弱。但是，在临床预测模型中加

入某炎性标记物所获得的临床外效用性的程度，不

仅仅是由多因素模型中的独立预测因素所决定。在

13


以临床变量为基础的模型中加入 IL-6 后 c 统计量稍

增加，使其不可能增加此模型临床有效的预测而不

需要抽血。我们没有对能最佳预测复发性血管事件

的 IL-6 水平设定阈值，因为这样会导致对标记物预

测能力的偏移性评价或导致研究结果无法重复 [10]。

诠释

以上数据表明，CRP 或 IL-6 的水平对于卒中后

复发性血管事件的发生并非是特异性的因果关系，

而观察到相关性更可能反映了动脉粥样硬化或其危

险因素，或尚不明确的诱发因素所产生的炎性反应。

与此相符的是，一些研究检测了功能性 CRP 和 IL-6的基因多态性 ( 在基线水平 CRP 或 IL-6 就存在差

异 )，发现其多态性并不与卒中 [11] 或其它闭塞性血

管事件 [12] 的风险增加相关。为明确人类的生理性炎

性反应与复发性血管事件之间的因果关系，必须通

过有明确抗炎效果，但对其它血管危险因素如胆固

醇、血压无直接作用的药物的随机临床试验证实。

推广性

我们发现在本卒中病人队列中，IL-6 水平与

复发性血管事件和死亡之间密切相关，这与近期基

于人群的前瞻性研究结果相一致 [13,14]。我们发现的

CRP、IL-6、纤维蛋白原与复发性卒中之间 [1,15,16]，

CRP、纤维蛋白原与死亡之间 [16,17] 的相关性结果与

之前的研究结果相一致 ( 详见补充材料表 I，网址 http://stroke.ahajournals.org)。

结论我们发现了在卒中后的患者中 IL-6、CRP 和纤

维蛋白原的高水平与闭塞性血管事件的发生率增高

相关。IL-6、CRP 和纤维蛋白原水平与卒中后血管

性死亡及非血管性死亡之间的相关性相似。今后，

在预测卒中后复发性血管事件的研究中，应采用简

单易测的临床变量或比较不同血标记物的作用。

参考文献1. Woodward M, Lowe GDO, Campbell DJ, Colman S, Rumley A, Chalmers

J, Neal BC, Patel A, Jenkins AJ, Kemp BE, MacMahon SW. Associations of inflammatory and hemostatic variables with the risk of recurrent stroke. Stroke.

2005;36:2143–2147.2. Welsh P, Lowe GDO, Chalmers J, Campbell DJ, Rumley A, Neal BC, MacMa-

hon SW, Woodward M. Associations of proinflammatory cytokines with the risk of recurrent stroke. Stroke. 2008;39:2226–2230.

3. Rothwell PM, Howard SC, Power DA, Gutnikov SA, Algra A, van GJ, Clark TG, Murphy MF, Warlow CP. Fibrinogen concentration and risk of ischemic stroke and acute coronary events in 5113 patients with transient ischemic attack and minor ischemic stroke. Stroke. 2004;35:2300–2305.

4. Grau AJ, Boddy AW, Dukovic DA, Buggle F, Lichy C, Brandt T, Hacke W, for the CAPRIE Investigators. Leukocyte count as an independent predictor of re-current ischemic events. Stroke. 2004;35:1147–1152.

5. Whiteley W, Jackson C, Lewis S, Lowe G, Rumley A, Sandercock P, Wardlaw J, Dennis M, Sudlow C. Inflammatory markers and poor outcome after stroke: a prospective cohort study and systematic review of interleukin-6. PLoS Med. 2009;6:e1000145.

6. Jackson CA, Hutchison A, Dennis MS, Wardlaw JM, Lewis SC, Sudlow CLM. Differences between ischemic stroke subtypes in vascular outcomes support a distinct lacunar ischemic stroke arteriopathy: a prospective, hospital-based study. Stroke. 2009;40:3679 –3684.

7. von Elm E, Altman DG, Egger M, Pocock SJ, Gotzsche PC, Vandenbroucke JP. The strengthening the reporting of observational studies in epidemiology (STROBE) statement: guidelines for reporting observational studies. Lancet. 2007;370:1453–1457.

8. Counsell C, Dennis M, McDowall M. Predicting functional outcome in acute stroke: comparison of a simple six variable model with other predictive systems and informal clinical prediction. J Neurol Neurosurg Psychiatry. 2004;75:401– 405.

9. Coviello V. Cumulative incidence estimation in the presence of competing risks. Stata J. 2004;4:103–112.

10. Royston P, Moons KGM, Altman DG, Vergouwe Y. Prognosis and prognostic research: developing a prognostic model. BMJ. 2009;338:b604.

11. Ladenvall C, Jood K, Blomstrand C, Nilsson S, Jern C, Ladenvall P. Serum C-reactive protein concentration and genotype in relation to ischemic stroke subtype. Stroke. 2006;37:2018 –2023.

12. Zacho J, Tybjaerg-Hansen A, Jensen JS, Grande P, Sillesen H, Nordestgaard BG. Genetically elevated C-reactive protein and ischemic vascular disease. N Engl J Med. 2008;359:1897–1908.

13. Danesh J, Kaptoge S, Mann AG, Sarwar N, Wood A, Angleman SB, Wensley F, Higgins JPT, Lennon L, Eiriksdottir G, Rumley A, Whincup PH, Lowe GDO, Gudnason V. Long-term interleukin-6 levels and subsequent risk of coronary heart disease: two new prospective studies and a systematic review. PLoS Med. 2008;5:e78.

14. Patterson CC, Smith AE, Yarnell JWG, Rumley A, Ben Shlomo Y, Lowe GDO. The associations of interleukin-6 (IL-6) and downstream inflammatory mark-ers with risk of cardiovascular disease: The Caerphilly Study. Atherosclerosis. 2010;209:551–557.

15. Campbell DJ, Woodward M, Chalmers JP, Colman SA, Jenkins AJ, Kemp BE, Neal BC, Patel A, MacMahon SW. Soluble vascular cell adhesion molecule 1 and N-terminal pro-B-type natriuretic peptide in predicting ischemic stroke in patients with cerebrovascular disease. Arch Neurol. 2006;63:60–65.

16. Elkind MS, Tai W, Coates K, Paik MC, Sacco RL. High-sensitivity C-reactive protein, lipoprotein-associated phospholipase A2, and outcome after ischemic stroke. Arch Intern Med. 2006;166:2073–2080.

17. Di Napoli M, Papa F, Bocola V. C-reactive protein in ischemic stroke: an inde-pendent prognostic factor. Stroke. 2001;32:917–924.

Association of Circulating Inflammatory Markers With...

Documents

Transcript of Association of Circulating Inflammatory Markers With...