relationship of coagulation and fibrinolytic variables with arterial structure and function in...

Thrombosis Research xxx (2014) xxx–xxx

TR-05499; No of Pages 6

Contents lists available at ScienceDirect

Thrombosis Research

j ourna l homepage: www.e lsev ie r .com/ locate / th romres

Regular Article

Relationship of coagulation and fibrinolytic variables with arterialstructure and function in Africans

M. Pieters a,⁎, H. Boshuizen b, Z. de Lange a, A.E. Schutte c, R. Schutte c, M. Greeff d, R.A.S. Ariëns e

a Centre of Excellence for Nutrition, North-West University, Potchefstroom, South Africab Department of Human Nutrition, Wageningen University and Research Centre, Wageningenc Hypertension in Africa Research Team (HART), North-West University, Potchefstroom, South Africad Africa Unit for Transdisciplinary Health Research, North-West University, Potchefstroom, South Africae Theme Thrombosis, Division of Cardiovascular and Diabetes Research, Multidisciplinary Cardiovascular Research Centre and Leeds Institute for Genetics, Health and Therapeutics, School ofMedicine, University of Leeds, UK

⁎ Corresponding author at: Centre of Excellence for NuPotchefstroom Campus, Private Bag X6001, Potchefstroo18 299 2462; fax: +27 18 299 2464.

E-mail address: [email protected] (M. Pieters

http://dx.doi.org/10.1016/j.thromres.2014.04.0210049-3848/© 2014 Elsevier Ltd. All rights reserved.

Please cite this article as: Pieters M, et al, ReThromb Res (2014), http://dx.doi.org/10.101

a b s t r a c t
a r t i c l e i n f o
Article history:
Received 19 March 2014Received in revised form 11 April 2014Accepted 21 April 2014Available online xxxx
Keywords:Arterial structure and functionCardiovascular diseaseDisease progressionFibrin clot lysis timeFibrinogen

Introduction:Although both coagulation andfibrinolysis are associatedwith cardiovascular disease (CVD) the un-derlying nature and pathways of many of these associations are still unclear. Our aimwas to determine which ofthe current or 5-year prior levels of total fibrinogen, fibrinogen γ’, plasminogen activator inhibitor-1 (PAI-1act)and global fibrinolytic potential were the stronger determinant of arterial structure and function.Materials and methods: This prospective study consisted of 2010 Africans over the age of 35 years with 5-yearfollow-up data available for 1288 participants. Cardiovascular measurements included arterial stiffness, bloodpressure and carotid intima media thickness.Results: Fibrinogenγ’ showed stronger associationswith blood pressure than totalfibrinogen also in the presenceof other CVD risk factors. PAI-1act was positively associated with blood pressure both cross-sectionally andprospectively, with the longitudinal association being the stronger determinant, also after adjustment forknown CVD risk factors. Clot lysis time (CLT) was positively associated, both prospectively and cross-
sectionally, with intima media thickness and negatively with markers of arterial stiffness but not after adjust-ment for known CVD risk factors.Conclusions: Fibrinogen γ’ was more strongly associated with CVD function than total fibrinogen. PAI-1act wassignificantly associated with blood pressure with changes in PAI-1 levels preceding changes in blood pressure.Different mechanisms may be at play determining arterial wall stiffness/thickening and blood pressure asobserved from the opposing associations with PAI-1act and CLT. CLT was not independently related to cardiovas-cular measures as its associations were weakened in the presence of other known CVD risk factors.
© 2014 Elsevier Ltd. All rights reserved.

Introduction

Haemostatic abnormalities such as hypercoagulability and hypo-fibrinolyis have been associated with thrombotic diseases such ascardiovascular diseases (CVD) [1]. The nature and pathways of these as-sociations is, however, still under investigation. Some of the coagulationvariables, such as fibrinogen, have been independently and consistentlylinked to the development of CVD in both cross-sectional and prospectivedata analysis [2]. Others, such as von Willebrand factor, FVII and FVIIIhave been associated with CVD only in cross-sectional analysis(e.g. case-control studies), or the association with future risk wassignificantly diminished after adjustment for other CVD risk factors,

trition, North-West University,m2520, South Africa. Tel.: +27

).

lationship of coagulation and6/j.thromres.2014.04.021

questioning its possible causal contribution to CVD development[2–5]. Also for fibrinogen there is discussion regarding its causalitysince Mendelian randomization studies have been inconclusive [6]and therapeutic lowering of fibrinogen has not shown benefit in studiesof ischemic stroke [7].

Although hypofibrinolysis is considered to be associated withthrombosis, existing prospective studies have not consistently shownan independent association between fibrinolytic factors and CVD.While most studies have found significant associations between plas-minogen activator inhibitor-1 (PAI-1), tissue plasminogen activatorantigen and/or activity (tPAag, tPAact) and plasminogen with severaltypes of CVD, these associations were significantly reduced in many ofthe studies after adjustment for other CVD risk factors [4,8–14]. One ofthe questions that are difficult to answer is whether changes in fibrino-lysis play a causal role in thrombosis or whether they reflect changesinduced by the ischemic insult. Some data is available investigatingthe relationship of fibrinolytic variables with CVD progression such as

fibrinolytic variables with arterial structure and function in Africans,

http://dx.doi.org/10.1016/j.thromres.2014.04.021

mailto:[email protected]


http://www.sciencedirect.com/science/journal/00493848


2 M. Pieters et al. / Thrombosis Research xxx (2014) xxx–xxx

vascular stiffness, hypertension and intima media thickness (IMT)[15–19] but the topic remains as of yet inconclusive. The relationshipbetween fibrinolysis and CVD can also be investigated through the useof assays reflecting global fibrinolytic potential (reported as clot lysistime, CLT, for example) instead of analyzing individual components ofthe fibrinolytic system. A limited number of studies exist that investi-gates the association between CLT and CVD events using different CLTassays [20–22] but no information is available regarding the associationof CLT with markers of arterial structure and function, which wouldaid in elucidating the mechanisms behind the associations with CVDevents.

Another haemostatic variable for which no prospective data exist,but which have been associated with CVD through case control studies,is fibrinogen γ’. Fibrinogen γ’ arises from a splice variant of the γ chainmRNA resulting from an alternative polyadenylation signal in intron 9[23]. Its levels have been associated with both venous and arterialthrombosis, but with opposite trends of lower levels in venous andhigher levels in arterial thrombosis [23]. Data collected in prospectivestudies investigating the association between haemostatic and fibrino-lytic variables and CVD have largely been obtained from white popula-tions. Results from the ARIC study showed that, while fibrinolysis playsa role in the early stages of atherosclerosis in white participants, thiswas not the case in African-Americans [15]. Studies investigating theseassociations in other ethnicities such as Africans, who are consideredan under-studied population in CVD etiology, are largely lacking.

The aim of our study was to investigate the association of selectedhaemostatic variables, including both coagulant and fibrinolytic pro-teins, with arterial structure and function in a prospective study includ-ing 2010 Africans. Total fibrinogen (the final substrate of coagulation),and its variantfibrinogenγ’(anovel, emerging risk factor)were selectedas coagulation variables with PAI-1act (the major inhibitor of clot lysis)and CLT (a global estimate of fibrinolytic potential) representingfibrinolytic variables. This was performed by investigating both thelongitudinal (prospective) as well as cross-sectional associations ofthese variables with markers of arterial structure and function such asvascular stiffness (augmentation index (AI) and pulse wave velocity),blood pressure and carotid intima media thickness.

Materials and methods

Study cohort

The Prospective Urban and Rural Epidemiological (PURE) study is alarge-scale cohort study that tracks changing lifestyles, risk factors andchronic disease in rural and urban areas of 21 countries from all majorcontinents over 12 years. For further details see Teo et al. [24]. InSouth Africa the baseline data was collected in 2005 and the firstfollow-up data in 2010. Two thousand and ten randomly selectedAfrican participants (1260 women and 750 men) living the NorthWest Province of South Africa participated in the PURE study in 2005.These participants were recruited from 6000 randomly selected house-holds from two communities based on representativeness and feasibil-ity for long-term follow-up, according to the guidelines stipulated in theoverarching PURE study [24,25]. Apparently healthy black SouthAfricans, older than 35, were eligible to participate. Exclusion criteriawere use of chronic medication for non-communicable diseases and/or any self-reported acute illness. In 2010, 1288 of the original 2010 par-ticipants were available for follow-up. Reasons for the smaller numberof participants were: refusal to participate again or withdrawal fromthe study, relocation, not being available on days of sample collectionor acute illness or death (n = 230). Efforts to retain participants andlimit loss during the five year period was made through three monthlyfollow-up visits by trained fieldworkers aswell as completion of annualfollow-up reports. The study complied with the Helsinki Declarationandwas approved by the ethics committee of the North-West Universi-ty. Subjects signed informed consent before taking part in the study. All

Please cite this article as: Pieters M, et al, Relationship of coagulation andThromb Res (2014), http://dx.doi.org/10.1016/j.thromres.2014.04.021

data were treated confidentially and all analyses were performed withcoded data with the operator performing laboratory assays blind topatient status.

Blood processing

Fasting blood samples were collected withminimum stasis from theantecubital veins of participants between 07:00 and 11:00 am. For theanalysis of fibrinogen γ’, total fibrinogen, PAI-1act and turbidimetricmeasurement of clot formation and lysis, blood was collected intocitrate tubes (1 part of 0.1 M Na citrate to 9 parts of blood). Sampleswere centrifuged at 2000 x g for 15 minutes at 10 °C within 30 minutesof collection. Aliquotswere frozen on dry ice, stored in thefield at -18 °Cand then after 2-4 days at -82 °C until analysis.

Analysis of coagulation and fibrinolysis parameters

Coagulation and fibrinolytic variables were analysed in 2005 and2010. Total fibrinogen concentrations were measured using a modifiedClauss method on an automated coagulation analyser (MultifibrinU-test, BCS analyser, Dade Behring, Deerfield, IL, USA for the 2005samples and ACL-200, Instrumentation Laboratories, Milan, Italy for2010). One hundred and forty of the 2005 samples were re-run withthe 2010 samples to exclude the possibility that batch differenceswere artefacts resulting from the use of the different analysers. Fibrino-gen γ’ was measured by ELISA using the 2.G.H9 mouse monoclonalcoating antibody against the human γ’ sequence from Santa CruzBiotechnology (Santa Cruz, USA) for antigen capture and a goat poly-clonal HRP-conjugated antibody against human fibrinogen fromAbcam for development (Cambridge, USA) [26,27]. PAI-1act was mea-sured using an indirect enzymatic method (Spectrolyze PAI-1, TrinityBiotech, Bray, Ireland). Plasma fibrinolytic potential of tissue factor in-duced clots, lysed by exogenous tPA was analyzed using turbidimetricanalysis (A405nm) according to the method of Lisman et al. [28], withslightlymodified tissue factor and tPA concentrations in order to obtainclot lysis times of about 60 min (intra-assay CV = 3.6%, between plateCV = 4.5%). Final concentrations were tissue factor (125 x diluted –

an estimated final concentration of 59 pM according to Duckers et al.[29]; Dade Innovin, Siemens, Marburg, Germany), CaCl2 (17 mmol/l),tPA (100 ng/ml; Actilyse, Boehringer Ingelheim, Ingelheim, Germany)and phospholipid vesicles (10 μmol/l; Rossix, Mölndal, Sweden). CLTwas defined as the time from the midpoint in the transition from theinitial baseline to maximum turbidity, which is representative of clotformation, to the midpoint in the transition from maximum turbidityto the final baseline turbidity, which represents the lysis of the clot[28]. The coefficient of variance for all assays was b 10%.

Cardiovascular measurements

The following cardiovascular measurements were performed at thefollow-up in 2010. Estimated central (cSBP) and brachial (bSBP) systolicblood pressure (also measured in 2005) and augmentation index (AI)weremeasuredwith participants seateduprightwith the right arm sup-ported at heart level, using theOmronHEM9000AI (OmronHealthCare,Kyoto, Japan). A SonoSite Micromax Ultrasound system (SonoSite Inc.,Bothell, WA) and 6-13MHz linear array transducer were used to deter-mine carotid intima-media thickness (cIMT). Images from at least twooptimal angles of the left and right common carotid artery were obtain-ed.Measurementswere conducted using the semi-automated program;Artery Measurement Systems II v1.139 (Chalmers University ofTechnology, Gothenburg, Sweden). Far wall measurements were used.Pulse wave velocity was measured using noninvasively accessiblesuperficial pulses and the Complior SP device (Artech-Medical, Pantin,France) in an upper limb muscular artery, over the carotid radial seg-ment (carotid-radial PWV [crPWV]), reflecting vascular stiffness inmuscular arteries and in an elastic muscular mixed arterial segment



3M. Pieters et al. / Thrombosis Research xxx (2014) xxx–xxx

over the carotid dorsalis pedis segment (carotid-dorsalis pedis PWV[cdPWV]), reflecting vascular stiffness in both conduit and musculararteries. The PWV was measured on the left side of each participant,while in a supine position.

Statistical analysis

Univariate linear regression was used to determine the cross-sectional association of the 2010 coagulation and fibrinolytic variableswith cardiovascular measures as well as to determine whether levelsof the coagulation and fibrinolytic variables in 2005may have a predic-tive contribution to the cardiovascular measures in 2010. Coagulationand fibrinolytic variables were entered as continuous variables and re-sults from these models reported as regression coefficients (β and 95%CI). For all analyses four models were constructed. The first twomodelscomprised of an individual coagulation or fibrinolytic variable (either2005 or 2010) with age and gender. In the third model the 2010 coagu-lation and fibrinolytic variables were entered together with known CVDrisk factors (age, gender, LDL-cholesterol, C-reactive protein, systolicblood pressure, HIV-status, body mass index, alcohol consumption,smoking and contraceptive use). In the fourthmodel, the 2005 coagula-tion or fibrinolytic variable was additionally added to determine itspossible longitudinal contribution to the outcome variables afteradjusting for the other variables in the model. For total and γ’ fibrino-gen, the analyses were repeated for individuals with a CRP ≤ 3 mg/Lin 2005 (n=891) to account for possible artificially high 2005fibrinogenlevels resulting froman acute phase response. To account formissing data(random occurrence of inadequate sample volume and/or haemolysis),multiple imputation (10x) was performed and results of the pooled anal-ysis reported. The computer software package SPSS® version 21 (PASWInc., Chicago, IL, USA) was used for statistical analyses. A p-value ≤ 0.05was regarded as statistically significant.

Results

The study population characteristics are presented in Table 1 of theOnline Supplement. Baseline characteristics of the total study population(n= 2010) did not differ from those available for follow-up (n= 1288).Detailed descriptions of the coagulation and fibrinolytic variables

Table 1Longitudinal and cross-sectional association of blood pressure with coagulation and fibrinolyti

Variable Model 1 Model 2

Variable 2005 + age, gender Variable 2010 + age, gend

β (95% CI) p β (95% CI) p

cSBPTotal fibrinogen 0.91 (-0.47; 2.30) 0.19 1.84 (-0.45; 4.14) 0

Fibrinogen γ’ 4.29 (-6.20; 14.8) 0.42 9.58 (-3.02; 22.2) 0

PAI-1act 0.39 (0.20; 0.57) b0.0001 0.26 (0.06; 0.47) 0

Clot lysis time 0.08 (-0.06; 0.22) 0.27 0.09 (-0.005; 0.22) 0

bSBPTotal fibrinogen 1.10 (-0.12; 2.32) 0.08 1.53 (-0.58; 3.65) 0

Fibrinogen γ’ 4.71 (-5.97; 15.4) 0.39 7.58 (-4.05; 19.2) 0

PAI-1act 0.38 (0.21; 0.56) b0.0001 0.28 (0.08; 0.47) 0

Clot lysis time 0.12 (0.0001; 0.23) 0.05 0.10 (-0.02; 0.23) 0

PAI-1act – plasminogen activator inhibitor-1 activity; β – regression coefficient; 95% CI – 95% cpressure.


analyzed in this population and their relationship with CVD risk factorshave been published elsewhere [27,30,31].

Coagulation: Total and γ’ fibrinogen

Total fibrinogen, and specifically fibrinogen γ’ in 2010 was signifi-cantly associatedwith cSBP also after adjustment for other CVD risk fac-tors (model 3: p=0.04 and 0.01,model 4: p=0.05 and 0.01) (Table 1).Fibrinogen γ’ in 2010 was also a significant contributor to bSBP evenafter adjustment for other CVD risk factors (model 3: p = 0.02; model4: p = 0.01). Total and γ’ fibrinogen level in 2005 did not contributelongitudinally to cSBP or bSBP. Total fibrinogen in 2010 was associatedwith IMT in the unadjusted model (Table 3). The relative amount of γ’(the γ’ ratio) did not indicate significantly different associations thanthe absolute γ’ concentration (data not reported).

Fibrinolysis: PAI-1act and CLT

Both PAI-1act in 2005 (p b 0.0001) and 2010 (p= 0.01)were signif-icantly associated with cSBP and after adjusting for the other CVD riskfactors, the 2005 PAI-1act level remained a significant predictor (model4: p = 0.02). Also for bSBP, both the 2005 and 2010 PAI-1act levelswere significant contributors, and the 2005 level again proved to bethe stronger predictor when combined in model 4 (p = 0.02 for PAI-1act in 2005 and p = 0.24 for PAI-1act in 2010) (Table 1). PAI-1act bothin 2005 and 2010 was significantly associated with cdPWV (OnlineSupplement Table 2), but after adjustment for other CVD risk factors,this association disappeared. Clot lysis time, both in 2005 and 2010,was negatively associated with two markers of arterial stiffness, AI(large arteries) and crPWV (muscular arteries) (Table 2) in the unad-justedmodels, but after adjustment for CVD risk factors, it was no longersignificantly associated. Clot lysis time in 2005 (p= 0.002) and 2010(p = 0.003) was, however, positively associated with IMT (Table 3).Again the association disappeared after adjustment for CVD risk factors.

The change in coagulation and fibrinolytic variables from 2005 to2010 (delta variables) did not provide additional information comparedto using the 2005 data. In some cases the delta variables were in agree-mentwith the 2005 data when significance was indicated, but in others

c variables.

Model 3 Model 4

er Variable 2010 + CVD riskfactors

Variable 2005 and 2010 + CVD riskfactors

β (95% CI) p β (95% CI) p

.12 1.66 (0.09; 3.22) 0.04 0.32 (-0.48;1.12) - 20051.57 (-0.02; 3.16) - 2010

0.430.05

.14 10.74 (2.47; 19.0) 0.01 -0.81 (-7.80; 6.20) - 200510.8 (2.50; 19.1) - 2010

0.820.01

.01 0.10 (-0.04; 0.23) 0.15 0.16 (0.03; 0.29) - 20050.05 (-0.09; 1.90) - 2010

0.020.51

.20 0.02 (-0.07; 0.11) 0.70 0.05 (-0.04; 0.15) - 20050.001 (-0.10; 0.10) - 2010

0.270.99

.16 1.00 (0.40; 1.6) 0.10 0.22 (-0.42; 0.86) - 20050.93 (-0.26; 2.13) - 2010

0.500.13

.20 7.94 (1.57; 14.3) 0.02 -2.34 (-7.66; 2.99) - 20058.21 (1.79; 14.6) - 2010

0.390.01

.005 0.11 (0.005; 0.22) 0.04 0.13 (0.02; 0.24) - 20050.07 (-0.05; 0.18) - 2010

0.020.24

.11 0.032 (-0.04; 0.11) 0.41 0.06 (-0.02; 0.13) - 20050.01 (-0.07; 0.09) - 2010

0.140.29

onfidence intervals. cSBP - central systolic blood pressure; bSBP - brachial systolic blood



Table 2Longitudinal and cross-sectional association of arterial stiffness indices of the large (AI) and muscular (crPWV) arteries with coagulation and fibrinolytic variables.

Variable Model 1 Model 2 Model 3 Model 4

Variable 2005, age, gender Variable 2010, age, gender Variable 2010, CVD risk factors Variable 2010 and 2005, CVD riskfactors

β (95% CI) p β (95% CI) p β (95% CI) p β (95% CI) p

AITotal fibrinogen 0.15 (-0.52; 0.82) 0.66 0.41 (-0.82; 1.64) 0.51 0.93 (-0.42; 2.28) 0.18 -0.06 (-0.79; 0.66) - 2005

0.95 (-0.42; 2.31) - 20100.860.17

Fibrinogen γ’ -4.86 (-9.97; 0.25) 0.06 4.03 (-2.66; 10.7) 0.24 3.79 (-3.42; 11.0) 0.30 -3.74 (-9.60; 2.11) - 20054.29 (-2.98; 11.6) - 2010

0.210.25

PAI-1act 0.04 (-0.06; 0.13) 0.43 0.04 (-0.07; 0.14) 0.48 0.08 (-0.02; 0.19) 0.12 0.07 (-0.03; 0.18) - 20050.06 (-0.05; 0.17) - 2010

0.160.30

Clot lysis time -0.13 (-0.19; -0.06) b0.0001 -0.13 (-0.20; -0.07) b0.0001 -0.02 (-0.09; 0.06) 0.63 -0.01 (-0.09; 0.07) - 2005-0.01 (-0.09; 0.07) - 2010

0.770.73

crPWVTotal fibrinogen 0.05 (-0.07; 0.18) 0.42 -0.15 (-0.39; 0.08) 0.20 -0.11 (-0.36; 0.15) 0.42 0.03 (-0.12; 0.17) - 2005

-0.11 (-0.37; 0.14) - 20100.720.39

Fibrinogen γ’ -0.67 (-1.84; 0.51) 0.26 -0.41 (-1.73; 0.92) 0.55 -0.59 (-1.98; 0.80) 0.40 -0.44 (-1.62; 0.75) - 2005-0.52 (-1.90; 0.87) - 2010

0.460.47

PAI-1act 0.001 (-0.02; 0.02) 0.91 -0.005 (-0.03; 0.02) 0.62 -0.004 (-0.02; 0.02) 0.71 0.009 (-0.001; 0.03) - 2005-0.007 (-0.03; 0.02) - 2010

0.400.54

Clot lysis time -0.01 (-0.03; -0.004) 0.009 -0.02 (-0.03; -0.006) 0.004 -0.005 (-0.02; 0.009) 0.46 0.01 (-0.004; 0.03) - 2005-0.009 (-0.02; 0.006) - 2010

0.160.24

PAI-1act – plasminogen activator inhibitor-1 activity; β – regression coefficient; 95% CI – 95% confidence intervals.


the deltas indicated no significance, while the 2005 value indicatedsignificance (data not reported).

Discussion

Although abnormalities in the haemostatic processes involved in co-agulation and fibrinolysis have been linked to thrombotic diseases, thepathways involved in many of these relationships remains to be re-solved. Especially in non-white ethnicities there is a paucity of availabledata, with emerging evidence from the ARIC study indicating that theseassociations may be different to what is observed in white individuals.This study provides evidence regarding the relationship of coagulationand fibrinolytic variableswithmarkers of arterial structure and functionin Africans in order to identify possible pathways through which thesevariables may be related to CVD. It sheds some light on the sequenceof events and whether current or 5 year prior haemostatic levels are as-sociated with cardiovascular measures at present. This is the first studyto investigate the prospective relationship of fibrinogen γ’with arterialstructure and function. It also, for the first time, employs a clot lysisassaywhich is considered to be a true reflection of globalfibrinolytic po-tential [28] to investigate the prospective association of lysis rate withCVD risk markers.

Table 3Longitudinal and cross-sectional association of cIMT with coagulation and fibrinolytic variables

Variable Model 1 Model 2

Variable 2005, age, gender Variable 2010, age, gender

β (95% CI) p β (95% CI) p

Total fibrinogen 0.0001 (-0.01; 0.009) 0.94 0.02 (0.002; 0.03) 0.02

Fibrinogen γ’ 0.06 (-0.02; 0.14) 0.15 0.04 (-0.04; 0.12) 0.28

PAI-1act 0.0001 (-0.001; 0.001) 0.73 0.001 (0.0001; 0.02) 0.22

Clot lysis time 0.001 (0.001; 0.002) 0.002 0.001 (0.0001; 0.002) 0.00

PAI-1act – plasminogen activator inhibitor-1 activity; β – regression coefficient; 95% CI – 95% c


Coagulation

Hypercoagulability is strongly associated with atherothromboticdiseases [1]. Fibrinogen and its variants, being the main substrates inthe coagulation pathway, are therefore considered to play an importantrole in this association [2]. This is the first study to provide evidence onthe association offibrinogenγ’with central and brachial blood pressure,both measures considered important predictors of future CVD risk. Ourresults indicate that it is a stronger contributor to both cSBP and bSBPthan total fibrinogen on a cross-sectional level. Five year prior levelswere not associated with any of the cardiovascular measures. Theabsence of a clear relationship between fibrinogen and most of the car-diovascular measures may be due to the fact that Africans are known tohave increased fibrinogen concentration even in the absence of CVD[32–34]. The reasons for these increased levels remain to be deter-mined. In a recent study [30] we investigated the possible contributionof environmental factors to fibrinogen concentration in this populationwhile a study investigating possible genetic influences is currentlyunderway. Fibrinogen is thought to contribute to vascular dysfunctionthrough several mechanisms. Binding of fibrinogen to its endothelialreceptors (such asαvβ3 andα5β1 integrins and ICAM-1)may activate sig-naling pathways that alter vasoactivity, e.g. causing vasoconstrictionwitha resulting increase in vascular tone, which compromise endothelial layer

.

Model 3 Model 4

Variable 2010, CVD risk factors Variable 2010 and 2005, CVD risk factors

β (95% CI) p β (95% CI) p

0.001 (-0.01; 0.02) 0.88 -0.002 (-0.01; 0.007) - 20050.002 (-0.01; 0.02) - 2010

0.680.84

-0.006 (-0.09; 0.08) 0.90 0.05 (-0.03; 0.13) - 2005-0.01 (-0.1; 0.07) - 2010

0.770.20

0.0001 (-0.001; 0.002) 0.70 -0.001 (-0.002; 0.001) - 20050.0001 (-0.001; 0.002) - 2010

0.420.54

3 -0.0001 (-0.001; 0.001) 0.92 0.0001 (-0.001; 0.001) - 2005-0.0001 (-0.001; 0.001) - 2010

0.800.86

onfidence intervals.



5M. Pieters et al. / Thrombosis Research xxx (2014) xxx–xxx

integrity resulting in fibrinogen deposition in the sub-endothelial matrix.It can also increase plasma viscosity and erythrocyte aggregation [35]. Inatherosclerotic plaque formation it promotes lipid deposition, attractsplatelets and macrophages and promotes smooth muscle cell prolifera-tion [36]. Previous studies investigating these associations, studied totalfibrinogen and did not investigate fibrinogen γ’ specifically. It remainsto be determined whether fibrinogen γ’ has altered binding propertiesfor endothelial membrane receptors compared to γA/γA fibrinogen orwhether the associations observed for total fibrinogen largely reflectproperties of fibrinogen γ’ specifically.

Fibrinolysis

In addition to hypercoagulability, hypofibrinolyis is also stronglylinked to atherothrombotic disease [1]. PAI-1act levels and CLT arefrequently used to represent fibrinolysis. PAI-1act level in 2005 was astronger contributor to blood pressure (both central and brachial pres-sure) than the 2010 level. These results indicate that PAI-1act may alsocontribute to the development of high blood pressure instead of beinga consequence thereof only. The observed association of PAI-1 withblood pressure traits is not new [16,19,37,38], although most of theevidence comes from studies with a cross-sectional design. It is postu-lated that hypertension can lead to elevated PAI-1 levels throughhypertension-induced shear stress and/or endothelial activation [39].We and others [18] however, found increased PAI-1 levels to associatewith blood pressure in prospective analysis. PAI-1 may contribute tostiffening of blood vessels through direct effects on the vessel wall, plas-min inhibition and inhibition of vascular smooth muscle cell migration[19]. Itmay furthermore contribute to the development of hypertensionin that overproduction of PAI-1 seems to accelerate perivascular andmedial fibrosis [40]. The relationship between arterial remodellingand hypertension is, however, complex in that vascular remodellingcontributes to both increased peripheral resistance and reduced elastic-ity therefore affecting both the development and complications ofhypertension [41]. In agreement with this, PAI-1act was not only associ-atedwith blood pressure in our study, but alsowith pulsewave velocity.

This is the first study to provide data on the prospective associationof CLTwith CVD disease progressionmarkers using an assay that is con-sidered to be a true reflection of an individual’s fibrinolytic potential[28]. Clot lysis time was negatively associated with markers of arterialstiffness (AI and crPWV)while positively associatedwith bSBP (border-line only) and cIMT. The positive associationwith bSBP is likely a reflec-tion of the strong association of PAI-1 with blood pressure. Its positiveassociation with cIMT indicates that increased CLT is already associatedwith early (sub-clinical) atherosclerosis. The inverse associations of CLTwith arterial stiffness (both longitudinal and cross-sectional) seemsomewhat counter-intuitive. Individual components of the fibrinolyticpathway have, however also been found to be positively associatedwith CVD. In a cross-sectional study not only tPAAg (likely reflectingPAI-1/tPA complexes) but also tPAact was found to be increased inearly atherosclerosis suggesting that tPA synthesis in the endotheliumis already activated in the early stages of atherosclerosis indicative ofendothelial dysfunction/injury [42,43]. Also in prospective studies, in-creased tPA and plasminogen levels preceded CVD events [8,11,12].Ridker et al. [44] concluded that activation of the fibrinolytic pathwaymay already occur years in advance of vascular occlusion. These resultsmay, however also be specific to our study population. In black SouthAfricans, stroke is far more prevalent than ischemic heart disease - theunderlying reason likely being the high prevalence of hypertension[45]. In our study population 47% of participants had hypertension. Ad-ditionally, hemorrhagic stroke is far more common in black SouthAfricans than in their white counterparts [45,46], with atherosclerosispresent in less than 1% of stroke patients in a study by Connor et al.[46]. Lastly Siegerink et al. [47] also found shorter CLT in stroke patientscompared to controls while CLT was increased in myocardial infarction.Since the associations of CLT with markers of disease progression did


not hold up in the presence of other CVD risk factors, it is likelythat CLT is not independently associated with these markers butmay reflect or mediate the relationship of other risk factors withCVD development.

A limitation of this study is that although a variety of haemostaticfactorswere included,we cannot exclude the possibility of residual con-founding. Also, although the prospective nature of the study allows us toinvestigate whether changes in coagulation and fibrinolytic variablesalready occur prior to a specific CVD disease progression state, it is notpossible to determine causality and while every attempt has beenmade to prevent possible selection bias, it is not impossible that itmay have occurred in some form.

In conclusion, our data provide evidence that in Africans, fibrinogenγ’ showed stronger associations with blood pressure than total fibrino-gen and that these associations held up in the presence of other CVD riskfactors. Five year prior PAI-1act wasmore strongly associatedwith bloodpressure than current PAI-1act levels indicating that changes in PAI-1levels precede changes in blood pressure. Different mechanisms maybe at play determining arterial wall stiffness, carotid wall thickeningand blood pressure as observed from the opposing associations withPAI-1act and CLT, suggesting that regulatory mechanism for CLT, otherthan PAI-1, can also significantly influence fibrinolytic potential. Thisfinding may, however be specific to our population which is known tohave an increased prevalence of hemorrhagic stroke. CLT does notseem to be independently related to CVD disease progression in thispopulation as its associations were weakened in the presence of otherknown CVD risk factors.

Authorship contribution

M. Pieters; performed research, collected data, analyzed andinterpreted data, performed statistical analysis and wrote manuscript.

H. Boshuizen; performed statistical analysis and critical reading ofmanuscript.

Z. de Lange; analyzed and interpreted data, performed laboratoryanalysis.

A.E. Schutte; performed research, collected data and criticallyreviewed manuscript.

R. Schutte; performed research, collected data and criticallyreviewed manuscript.

M. Greeff; designed research, collected data.R.A.S.Ariëns; analyzed and interpreted data and critically reviewed

the manuscript.

Funding

This work was supported by the Medical Research Council of SouthAfrica (MP), the South African Sugar Association (Project 214) (MP)and the South African National Research Foundation (NRF GUNnumbers 2069139, FA2006040700010 and FA2006041100003). Anyopinion, finding and conclusion or recommendation expressed in thismaterial is that of the author(s) and theNRFdoes not accept any liabilityin this regard. RASA is supported by the Medical Research Council(G0901546) and the British Heart Foundation (RG/13/2/30104).

Conflict-of-interest disclosure

None of the authors have any conflict of interest to declare.

Acknowledgements

We thank the PURE-SA research team, field-workers and partici-pants and the PURE project staff at the PHRI, Hamilton Health Sciencesand McMaster University, ON, Canada.




Appendix A. Supplementary data

Supplementary data to this article can be found online at http://dx.doi.org/10.1016/j.thromres.2014.04.021.

References

[1] Ajjan R, Grant PJ. Coagulation and atherothrombotic disease. Atherosclerosis2006;186:240–59.

[2] Danesh J, Lewington S, Thompson SG, Lowe GDO, Collins R. Plasma fibrinogen leveland the risk of major cardiovascular diseases and nonvascular mortality: an individualparticipant meta-analysis. JAMA 2005;294:1799–809.

[3] Kannel WB. Overview of hemostatic factors involved in atherosclerotic cardiovasculardisease. Lipids 2005;40:1215–20.

[4] Tzoulaki I, Murray GD, Lee AJ, Rumley A, Lowe GD, Fowkes FG. Relative value ofinflammatory, hemostatic, and rheological factors for incident myocardial infarctionand stroke: the Edinburgh Artery Study. Circulation 2007;115:2119–27.

[5] Willeit P, Thompson A, Aspelund T, Rumley A, Eiriksdottir G, Lowe G, et al. Hemo-static factors and risk of coronary heart disease in general populations: newprospective study and updated meta-analyses. PLoS One 2013;8:e55175.

[6] Van der Bom JG, de Maat MP, Bots ML, Haverkate F, de Jong PT, Hofman A, et al.Elevated plasma fibrinogen: cause or consequence of cardiovascular disease?Arterioscler Thromb Vasc Biol 1998;18:621–5.

[7] Hennerici MG, Kay R, Bogousslavsky J, Lenzi GL, Verstraete M, Orgogozo JM. Intrave-nous ancrod for acute ischaemic stroke in the European Stroke Treatment withAncrod Trial: a randomised controlled trial. Lancet 2006;368:1871–8.

[8] Thogersen AM, Jansson JH, Boman K, Nilson TK, Weinehall L, Huhtasaari F, et al. Highplasminogen activator inhibitor and tissue plasminogen activator levels in plasma pre-cede a first acutemyocardial infarction in bothmen andwomen: evidence for thefibri-nolytic system as an independent primary risk factor. Circulation 1998;98:2241–7.

[9] Cushman M, Lemaitre RN, Kuller LH, Psaty BM, Macy EM, Sharrett AR, et al. Fibrino-lytic activation markers predict myocardial infarction in the elderly. The Cardiovas-cular Health Study. Arterioscler Thromb Vasc Biol 1999;19:493–8.

[10] Johansson L, Jansson JH, BomanK, Nilsson TK, Stegmayr B, Hallmans G. Tissue plasmin-ogen activator, plasminogen activator inhibitor-1, and tissue plasminogen activator/plasminogen activator inhibitor-1 complex as risk factors for the development of afirst stroke. Stroke 2000;31:26–32.

[11] Folsom AR, Aleksic N, Park E, Salomaa V, Juneja H, Wu KK. Prospective study offibrinolytic factors and incident coronary heart disease: the Atherosclerosis Risk inCommunities (ARIC) Study. Arterioscler Thromb Vasc Biol 2001;21:611–7.

[12] Lowe GD, Danesh J, Lewington S, Walker M, Lennon L, Thomson A, et al. Tissueplasminogen activator antigen and coronary heart disease. Prospective study andmeta-analysis. Eur Heart J 2004;25:252–9.

[13] Tzoulaki I, Murray GD, Lee AJ, Rumley A, Lowe GD, Fowkes FG. Inflammatory,haemostatic, and rheological markers for incident peripheral arterial disease:Edinburgh Artery Study. Eur Heart J 2007;28:354–62.

[14] Tzoulaki I, Murray GD, Price JF, Smith FB, Lee AJ, Rumley A, et al. Hemostatic factors,inflammatory markers, and progressive peripheral atherosclerosis: the EdinburghArtery Study. Am J Epidemiol 2006;163:334–41.

[15] SalomaaV, Stinson V, Kark JD, FolsomAR, Davis CE,WuKK. Association offibrinolyticparameters with early atherosclerosis. The ARIC Study. Atherosclerosis Risk inCommunities Study. Circulation 1995;91:284–90.

[16] Nishiwaki Y, Takebayashi T, Omae K, Ishizuka C, Nomiyama T, Sakurai H. Relation-ship between the blood coagulation-fibrinolysis system and the subclinical indica-tors of arteriosclerosis in a healthy male population. J Epidemiol 2000;10:34–41.

[17] Luttun A, Lupu F, Storkebaum E, Hoylaerts MF, Moons L, Crawley J, et al. Lack of plas-minogen activator inhibitor-1 promotes growth and abnormal matrix remodeling ofadvanced atherosclerotic plaques in apolipoprotein E-deficient mice. ArteriosclerThromb Vasc Biol 2002;22:499–505.

[18] Wang TJ, Gona P, Larson MG, Levy D, Benjamin EJ, Tofler GH, et al. Multiplebiomarkers and the risk of incident hypertension. Hypertension 2007;49:432–8.

[19] Lieb W, Larson MG, Benjamin EJ, Yin X, Tofler GH, Selhub J, et al. Multimarkerapproach to evaluate correlates of vascular stiffness: the Framingham Heart Study.Circulation 2009;119:37–43.

[20] Traby L, Kollars M, Eischer L, Eichinger S, Kyrle PA. Prediction of recurrent venousthromboembolism by clot lysis time: a prospective cohort study. PLoS One 2012;7:e51447.

[21] Meade TW, Cooper JA, Chakrabarti R, Miller GJ, Stirling Y, Howarth DJ. Fibrinolyticactivity and clotting factors in ischaemic heart disease in women. BMJ 1996;312:1581.

[22] Meade TW, Ruddock V, Stirling Y, Chakrabarti R, Miller GJ. Fibrinolytic activity,clotting factors, and long-term incidence of ischaemic heart disease in theNorthwick Park Heart Study. Lancet 1993;342:1076–9.


[23] Uitte de Willige S, Standeven KF, Philippou H, Ariens RA. The pleiotropic role of thefibrinogen gamma' chain in hemostasis. Blood 2009;114:3994–4001.

[24] Teo K, Chow CK, VazM, Rangarajan S, Yusuf S. The Prospective Urban Rural Epidemi-ology (PURE) study: examining the impact of societal influences on chronic non-communicable diseases in low-, middle-, and high-income countries. Am Heart J2009;158:1–7.

[25] Yusuf S, Islam S, Chow CK, Rangarajan S, Dagenais G, Diaz R, et al. Use of secondaryprevention drugs for cardiovascular disease in the community in high-income,middle-income, and low-income countries (the PURE Study): a prospective epide-miological survey. Lancet 2011;378:1231–43.

[26] Uitte de Willige S, de Visser MC, Houwing-Duistermaat JJ, Rosendaal FR, Vos HL,Bertina RM. Genetic variation in the fibrinogen gamma gene increases the risk fordeep venous thrombosis by reducing plasma fibrinogen gamma' levels. Blood2005;106:4176–83.

[27] Pieters M, Kotze RC, Jerling JC, Kruger A, Ariens RA. Evidence that fibrinogen gamma'regulates plasma clot structure and lysis and relationship to cardiovascular riskfactors in black Africans. Blood 2013;121:3254–60.

[28] LismanT, deGroot PG,Meijers JC, Rosendaal FR. Reducedplasmafibrinolytic potential isa risk factor for venous thrombosis. Blood 2005;105:1102–5.

[29] Duckers C, Simioni P, Spiezia L, Radu C, Dabrilli P, Gavasso S, et al. Residual plateletfactor V ensures thrombin generation in patients with severe congenital factor Vdeficiency and mild bleeding symptoms. Blood 2010;115:879–86.

[30] PietersM, deMaatMP, Jerling JC, Hoekstra T, Kruger A. Fibrinogen concentration andits role in CVD risk in black South Africans–effect of urbanisation. Thromb Haemost2011;106:448–56.

[31] De Lange Z, Pieters M, Jerling JC, Kruger A, Rijken DC. Plasma clot lysis time and itsassociation with cardiovascular risk factors in black Africans. PLoS One 2012;7:e48881.

[32] Pieters M, Vorster HH. Nutrition and hemostasis: a focus on urbanization in SouthAfrica. Mol Nutr Food Res 2008;52:164–72.

[33] Kaptoge S, White IR, Thompson SG, Wood AM, Lewington S, Lowe GDO, et al.Associations of plasma fibrinogen levels with established cardiovascular diseaserisk factors, inflammatory markers, and other characteristics: individual participantmeta-analysis of 154,211 adults in 31 prospective studies: the fibrinogen studiescollaboration. Am J Epidemiol 2007;166:867–79.

[34] Folsom AR, Wu KK, Conlan MG, Finch A, Davis CE, Marcucci G, et al. Distributions ofhemostatic variables in blacks and whites: population reference values from theAtherosclerosis Risk in Communities (ARIC) Study. Ethn Dis 1992;2:35–46.

[35] Lominadze D, Dean WL, Tyagi SC, Roberts AM. Mechanisms of fibrinogen-inducedmicrovascular dysfunction during cardiovascular disease. Acta Physiol (Oxf)2010;198:1–13.

[36] deMoerloose P, Boehlen F, Neerman-ArbezM. Fibrinogen and the risk of thrombosis.Semin Thromb Hemost 2010;36:7–17.

[37] Wall U, Jern C, Bergbrant A, Jern S. Enhanced levels of tissue-type plasminogenactivator in borderline hypertension. Hypertension 1995;26:796–800.

[38] Poli KA, Tofler GH, Larson MG, Evans JC, Sutherland PA, Lipinska I, et al. Associationof blood pressure with fibrinolytic potential in the Framingham offspring popula-tion. Circulation 2000;101:264–9.

[39] Lip GY, Blann AD. Does hypertension confer a prothrombotic state? Virchow's triadrevisited. Circulation 2000;101:218–20.

[40] Kaikita K, Fogo AB, Ma L, Schoenhard JA, Brown NJ, Vaughan DE. Plasminogenactivator inhibitor-1 deficiency prevents hypertension and vascular fibrosis in re-sponse to long-term nitric oxide synthase inhibition. Circulation 2001;104:839–44.

[41] Intengan HD, Schiffrin EL. Vascular remodeling in hypertension: roles of apoptosis,inflammation, and fibrosis. Hypertension 2001;38:581–7.

[42] SalomaaV, Stinson V, Kark JD, FolsomAR, Davis CE,WuKK. Association offibrinolyticparameters with early atherosclerosis. The ARIC Study. Atherosclerosis Risk inCommunities Study. Circulation 1995;91:284–90.

[43] Tomiyama H, Kimura Y, Mitsuhashi H, Kinouchi T, Yoshida H, Kushiro T, et al.Relationship between endothelial function and fibrinolysis in early hypertension.Hypertension 1998;31:321–7.

[44] Ridker PM, Hennekens CH, Stampfer MJ, Manson JE, Vaughan DE. Prospective studyof endogenous tissue plasminogen activator and risk of stroke. Lancet 1994;343:940–3.

[45] Mbewu A, Mbanya JC. Cardiovascular Disease; 2006.[46] Connor MD, Modi G, Warlow CP. Differences in the nature of stroke in a multiethnic

urban South African population: the Johannesburg hospital stroke register. Stroke2009;40:355–62.

[47] Siegerink B,MeltzerME, DeGroot PG, Algra A, Lisman T, Rosendaal FR. Clot lysis timeand the risk of myocardial infarction and ischaemic stroke in young women; resultsfrom the RATIO case-control study. Br J Haematol 2012;156:252–8.




http://refhub.elsevier.com/S0049-3848(14)00235-7/rf0005










































































































































relationship of coagulation and fibrinolytic variables with arterial structure and function in...

Documents