Search tips
Search criteria 


Logo of heartHeartVisit this articleSubmit a manuscriptReceive email alertsContact usBMJ
Heart. 2007 April; 93(4): 495–499.
Published online 2006 October 3. doi:  10.1136/hrt.2006.099374
PMCID: PMC1861468

Effects of blood pressure on the prothrombotic risk in 1235 patients with non‐valvular atrial fibrillation



Increased levels of plasma von Willebrand factor (vWf, an index of endothelial damage/dysfunction) and soluble P‐selectin (sP‐sel, an index of platelet activation) concentrations have been reported as indices of the prothrombotic state in both non‐valvular atrial fibrillation and hypertension separately. However, the effect of hypertension on the levels of these indices in the setting of atrial fibrillation, and whether increasing severity of hypertension presents an additive prothrombotic risk, is unclear.


Plasma concentrations of vWf and sP‐sel were measured by ELISA in 1235 patients with atrial fibrillation, and levels related to a history of hypertension and rising quartiles of systolic, diastolic and pulse pressure in those with and without diabetes mellitus and prior vascular events.


Mean plasma vWf was higher among patients with atrial fibrillation with a history of hypertension (149 vs 145 IU/dl, p = 0.005). Also, an increase in the levels of vWf with increasing quartiles of pulse pressure (p = 0.042) was noticed. However, on multivariate analysis, after adjusting for potential confounders, the effects of both hypertension and pulse pressure became non‐significant (p = 0.261 and p = 0.5, respectively). Levels of sP‐sel were unaffected by a history of hypertension and rising quartiles of systolic and diastolic blood pressure, or pulse pressure.


Among patients with atrial fibrillation, patients with hypertension have higher vWf levels, indicating endothelial damage/dysfunction, which is associated with increasing pulse pressure. However, these associations are probably owing to the presence of other associated cardiovascular disease, rather than hypertension itself. Furthermore, platelet activation (sP‐sel) was unrelated to hypertension or blood pressure in this atrial fibrillation cohort. Hypertension or blood pressure levels do not seem to have an independent additive affect on the prothrombotic state in atrial fibrillation.

Hypertension is a well‐recognised risk factor for coronary artery disease, atherosclerosis and cerebrovascular disease. In hypertension, the arterial walls are exposed to the surge of blood under high pressures; yet, the complications of hypertension, such as myocardial infarction or stroke, are paradoxically thrombotic rather than haemorrhagic—the so‐called “thrombotic paradox of hypertension”.1 Indeed, hypertension fulfils the components of the Virchow's triad,2 which includes abnormalities of the vessel wall (which we now recognise as endothelial damage/dysfunction), abnormal blood constituents (with abnormal levels of haemostatic, fibrinolytic and platelet factors) and abnormalities of blood flow.3

Non‐valvular atrial fibrillation is the most common sustained arrhythmia, and is also recognised to increase the risk of stroke by fivefold.4 However, the exact mechanisms of thromboembolism in the setting of atrial fibrillation remain incompletely understood.5 Abnormal plasma markers of coagulation, endothelial function and platelet activation have been described in the setting of atrial fibrillation, in comparison with healthy controls.6 For example, increased plasma levels of von Willebrand factor (vWf, a marker of endothelial damage/dysfunction) have been shown to be prognostically significant in atrial fibrillation,7 whereas increased plasma levels of soluble P‐selectin (sP‐sel, a marker of platelet activation) have been related to risk factors for atherosclerosis in patients with atrial fibrillation.8 However, many common clinical features can also influence the risk of stroke and thromboembolism in atrial fibrillation.4 Of note, atrial fibrillation is common among patients with hypertension and constitutes part of the spectrum of hypertensive heart disease, and a combination of presence of the atrial fibrillation and hypertension would raise the thrombotic risk of stroke in this population.

Indeed, hypertension itself is manifest with an array of aberrant inflammatory9 and metabolic factors10 that may underline the prothrombotic risk associated with atrial fibrillation. We have previously shown that vWf levels increase with additional risk factors contributing to the metabolic syndrome in hypertension,10 implying a relationship of increasing vWf levels with a greater cardiovascular risk “burden”. Thus, we hypothesised that the presence of coexistent hypertension and increasing severity of blood pressure (and pulse pressure) would have an additive and independent effect on the prothrombotic risk in atrial fibrillation. We therefore examined the effects of blood pressure on the prothrombotic risk, as reflected by levels of vWf and sP‐sel, in a large cohort of patients with non‐valvular atrial fibrillation.

Patients and methods

We studied 1235 consecutive patients with non‐valvular atrial fibrillation who were participants in the 3rd Stroke Prevention in Atrial Fibrillation study (SPAF‐III). We included all SPAF patients in this analysis, irrespective of antithrombotic treatment use, as warfarin or aspirin did not considerably influence vWf or sP‐sel levels in this cohort8 or in other studies.11 Patients with documented congestive heart failure were excluded from the study, as congestive heart failure has already been reported to increase prothrombotic risk in atrial fibrillation.12 The remaining patients with atrial fibrillation were divided into two groups: those with hypertension (defined as untreated systolic blood pressure (SBP) >160 mm Hg and diastolic blood pressure (DBP) >90 mm Hg, or a history of hypertension requiring antihypertensive drug treatment) and those without hypertension.

Blood collection and laboratory analysis

Blood for vWf and sP‐sel assays was drawn into 3.8% sodium citrate tubes (Becton Dickinson, BD Biosciences, Oxford, UK), immediately mixed by gentle inversion, stored on melting ice, centrifuged at 4°C for 30 000 g‐min within 1 h of phlebotomy, and plasma was separated for vWf and sP‐sel assays.

Measurements of sP‐sel and vWf were performed using ELISA with reagents from R&D Systems (Abington, UK) and Dako‐Patts (Ely, UK), respectively. The unit for vWf is IU/dl and was standardised by reference vWf from the National Institute for Biological Standards and Controls, Blanche Lane, South Mimms, Potters Bar, Hertfordshire, UK. Intraassay coefficients of variation for all ELISA assays were <5%, inter‐assay variances were <10%.

Data analysis

We initially compared the cohorts on the basis of the presence or absence of hypertension. As prior existence of vascular disease entails the presence of damage/endothelial dysfunction, we performed additional analyses after excluding patients with prior stroke, coronary artery bypass surgery, peripheral vascular disease, diabetes, myocardial infarction and angina, and diabetes mellitus. Finally, we divided this cohort into quartiles of systolic, diastolic and pulse pressure to explore a relationship between rising quartiles and either vWf or sP‐sel levels, if any.

Normally distributed data are expressed as mean (standard deviation (SD)). Differences in distribution of variables between groups were evaluated using two‐sample t tests and one‐way analysis of variance, as appropriate, with Tukey's post hoc test for intergroup comparisons. The non‐parametric distribution of vWf and sP‐sel in this population was determined using normality testing and expressed as median (interquartile range). Differences in non‐parametric variables were evaluated with the Mann Whitney U test and Kruskal Wallis test. Forward and backward stepwise linear regression analyses were used in multivariate analysis. Statistical significance was accepted at the 0.05 level (two‐sided). Statistical analyses were undertaken using SPSS software.


In the whole cohort (n = 1235), vWf levels were significantly higher (p = 0.005), but sP‐sel was not significantly different (p = 0.526), when patients with hypertension were compared with those without hypertension (table 11).). As expected, significant differences were found between the two groups with respect to the presence of vascular disease.

Table thumbnail
Table 1 Baseline clinical characteristics of 1235 patients with non‐valvular atrial fibrillation

When the patients were divided into quartiles on the basis of mean SBP, DBP and pulse pressure, no significant ordinal relationships were seen between vWf and sP‐sel levels and rising quartiles of mean SBP and DBP (table 22).). However, vWf (p = 0.042), but not sP‐sel (p = 0.45), was found to be significantly associated with rising quartiles of pulse pressure in the whole cohort (table 22).

Table thumbnail
Table 2 Relationship of plasma von Willebrand factor and soluble P‐selectin levels of quartiles of systolic, diastolic and pulse pressure (n = 1235)

After excluding patients with diabetes or prior vascular events, vWf and sP‐sel levels were not significantly related to blood pressure levels (table 33).). On a similar analysis of quartiles of SBP, DBP and pulse pressure, no significant ordinal relationships were seen between vWf and sP‐sel levels and rising quartiles of SBP, DBP or pulse pressure (table 44).). Thus, after adjusting for potential confounders, the effect of both hypertension and pulse pressure on vWf became non‐significant.

Table thumbnail
Table 3 Baseline clinical characteristics of 664 patients with non‐valvular atrial fibrillation but no diabetes or prior vascular events
Table thumbnail
Table 4 Relationship of plasma von Willebrand factor and soluble P‐selectin levels with quartiles of systolic, diastolic and pulse pressure (n = 664)


Hypertension is a strong clinical risk factor for stroke and atrial fibrillation, although many hypertensives often have comorbidities (eg, diabetes, vascular disease, etc) which are themselves associated with stroke and atrial fibrillation, and may confound any relationship. In our cohort of patients with non‐valvular atrial fibrillation, although blood pressure seemed to influence vWf levels, consistent with current thrombotic risk stratification regimens,4 subanalyses of groups after excluding those with diabetes or prior vascular disease showed that the relationship with vWf was not significant.

vWf is one of several endothelium‐derived haemostatic mediators, with key roles in platelet aggregation and stabilisation of circulating clotting factors.13 Large quantities of vWf are stored in the Weibel–Palade bodies of endothelial cells and can be mobilised rapidly after endothelial cell activation13 to result in transient elevations of plasma vWf.14 In experimental models at least, endothelial dysfunction promotes thrombosis, vasospasm and vessel occlusion.15 Thus, persistently raised levels of vWf and endothelial damage/dysfunction may pose a real risk to thrombosis. Moreover, raised vWf has been linked to the presence of left atrial appendage thrombus (as detected by transoesophageal echocardiography),16 and indeed there are reports showing a significant correlation between the degree of endocardial expression of vWf and the degree of platelet adhesion/thrombus formation in the atrial appendage.17 Of note, raised plasma vWf has also been shown to be present in patients with damaged atrial appendage endocardium, and also mitral valve disease, many of whom had atrial fibrillation.18 vWf has also been found to be a significant independent predictor of target organ damage in hypertension19 and to correlate well with Framingham risk‐factor prediction scores.20 In this study, mean plasma vWf was significantly higher among patients with atrial fibrillation with a history of hypertension and with increasing quartiles of pulse pressure. The increasing quartiles of pulse pressure can be related to vascular dysfunction, hypertensive target organ damage and prognosis.21,22 Interestingly, after adjusting for potential confounders, the effects of both hypertension and pulse pressure became non‐significant, suggesting that hypertension or blood pressure levels may not have an independent additive affect on the prothrombotic state in atrial fibrillation. Indeed, the prothrombotic state in atrial fibrillation may be more driven by concomitant comorbidities and vascular disease, and well‐controlled hypertension may perhaps represent less of a risk factor for stroke and vascular events in atrial fibrillation.23

Of note, sP‐sel were unaffected by a history of hypertension and rising quartiles of SBP and DBP, or pulse pressure, in this analysis. P‐selectin is a component of platelet α‐granules that is expressed on the platelet surface membrane and shed into the plasma (as sP‐sel) on platelet activation.24 P‐selectin is also found in Weibel–Palade bodies of endothelial cells, but the bulk of circulating sP‐sel seems to be platelet derived.25,26 Although vWf is similarly found in platelet α‐granules and endothelial cell Weibel–Palade bodies, most of the circulating vWf seems to be derived from the endothelium.14 The lack of relationship of sP‐sel and hypertension severity is consistent with observations that platelet activation—although present in atrial fibrillation—is more related to associated comorbidities, rather than to atrial fibrillation in itself, where coagulation factor‐related abnormalities predominate in contributing to the prothrombotic state.27 This is evident by the lack of relationship of sP‐sel with prognosis,7 stroke risk stratification,8 and the superior benefits of warfarin over antiplatelet treatment.28

This study is limited by its cross‐sectional design. Furthermore, we have limited information on concomitant drug treatments, and some antihypertensive drugs (eg, angiotensin‐converting enzyme inhibitors) may affect the endothelium and platelets. Also, a recent small study suggests that digoxin may influence flow cytometric assessments of platelet and endothelial activation,28 but we are unaware of digoxin affecting vWf and sP‐sel levels. Some patients were also taking warfarin and aspirin at baseline, but previous work does not suggest a significant influence on vWf and sP‐sel levels.8,11,29,30

Although abnormally high levels of circulating markers of endothelial function and platelet activation have been described in patients with non‐valvular atrial fibrillation,6 this study shows the lack of an independent effect of blood pressure on these markers in the same setting. This is in keeping with our earlier observations from a smaller cohort of patients with atrial fibrillation31 as well as a multivariate analysis, where neither hypertension nor SBP exceeding 160 mm Hg were found to be independently correlated with increased plasma concentration of either vWF or sP‐sel.8 This may perhaps be explained by the fact that concentrations of vWf may depend more strongly on other factors in the hypertensive microenvironment (eg, inflammation, angiogenesis, etc) rather than on the absolute level of blood pressure in itself.9,32,33 Thus, hypertension or the severity of blood pressure elevation may not have an independent additive affect on the prothrombotic state in atrial fibrillation beyond that seen with concomitant vascular disease and comorbidities associated with atrial fibrillation.


We thank the support of the Sandwell and West Birmingham Hospitals NHS Trust Research and Development programme for the Haemostasis, Thrombosis and Vascular Biology Unit for their support. We also thank Dr Robert G Hart for his helpful advice, and Dr D Conway and Dr B Chin for assistance with data collection.


DBP - diastolic blood pressure

SBP - systolic blood pressure

sP‐sel - soluble P‐selectin

vWf - plasma von Willebrand factor


Competing interests: None declared.


1. Lip G Y H. Hypertension, platelets, and the endothelium: the “thrombotic paradox” of hypertension (or “Birmingham paradox”) revisited. Hypertension 2003. 41199–200.200 [PubMed]
2. Virchow R. Phlogose und Thrombose in Gerasssystem. In: Virchow R, ed. Gesammelte Abhandlungen zur Wissenchaftichen Medicin. Frankfurt, Germany,Von Meidinger Sohn 1856. 458–636.636
3. Varughese G I, Lip G Y. Is hypertension a prothrombotic state? Curr Hypertens Rep 2005. 7168–173.173 [PubMed]
4. Lip G Y, Boos C. Antithrombotic therapy for atrial fibrillation. Heart 2006. 92155–161.161 [PMC free article] [PubMed]
5. Hart R G, Halperin J L. Atrial fibrillation and stroke: concepts and controversies. Stroke 2001. 32803–808.808 [PubMed]
6. Choudhury A, Lip G Y. Atrial fibrillation and the hypercoagulable state: from basic science to clinical practice. Pathophysiol Haemost Thromb 2003–2004. 33282–289.289 [PubMed]
7. Conway D S, Pearce L A, Chin B S. et al Prognostic value of plasma von Willebrand factor and soluble P‐selectin as indices of endothelial damage and platelet activation in 994 patients with nonvalvular atrial fibrillation. Circulation 2003. 1073141–3145.3145 [PubMed]
8. Conway D S, Pearce L A, Chin B S. et al Plasma von Willebrand factor and soluble p‐selectin as indices of endothelial damage and platelet activation in 1321 patients with nonvalvular atrial fibrillation: relationship to stroke risk factors. Circulation 2002. 1061962–1967.1967 [PubMed]
9. Patel J V, Lim H S, Nadar S. et al Abnormal soluble CD40 ligand and C‐reactive protein concentrations in hypertension: relationship to indices of angiogenesis. J Hypertens 2006. 24117–121.121 [PubMed]
10. Lim H S, Lip G Y, Blann A D. Plasma von Willebrand factor and the development of the metabolic syndrome in patients with hypertension. J Clin Endocrinol Metab 2004. 895377–5381.5381 [PubMed]
11. Lip G Y, Lowe G D, Rumley A. et al Increased markers of thrombogenesis in chronic atrial fibrillation: effects of warfarin treatment. Br Heart J 1995. 73527–533.533 [PMC free article] [PubMed]
12. Lip G Y, Pearce L A, Chin B S. et al Effects of congestive heart failure on plasma von Willebrand factor and soluble P–selectin concentrations in patients with non‐valvar atrial fibrillation. Heart 2005. 91759–763.763 [PMC free article] [PubMed]
13. Wagner D D. Cell biology of von Willebrand factor. Annu Rev Cell Biol 1990. 6217–246.246 [PubMed]
14. Lip G Y, Blann A D. von Wilebrand factor: a marker of endothelial dysfunction in vascular disorders? Cardiovasc Res 1997. 34255–265.265 [PubMed]
15. Luscher T F. The endothelium and cardiovascular disease–a complex relation. N Engl J Med 1994. 3301081–1083.1083 [PubMed]
16. Heppell R M, Berkin K E, McLenachan J M. et al Haemostatic and haemodynamic abnormalities associated with left atrial thrombosis in non‐rheumatic atrial fibrillation. Heart 1997. 77407–411.411 [PMC free article] [PubMed]
17. Fukuchi M, Watanabe J, Kumagai K. et al Increased von Willebrand factor in the endocardium as a local predisposing factor for thrombogenesis in overloaded human atrial appendage. J Am Coll Cardiol 2001. 371436–1442.1442 [PubMed]
18. Goldsmith I, Kumar P, Carter P. et al Atrial endocardial changes in mitral valve disease: a scanning electron microscopy study. Am Heart J 2000. 140777–784.784 [PubMed]
19. Spencer C G, Gurney D, Blann A D. et al Von Willebrand factor, soluble P‐selectin, and target organ damage in hypertension: a substudy of the Anglo‐Scandinavian Cardiac Outcomes Trial (ASCOT). Hypertension 2002. 4061–66.66 [PubMed]
20. Blann A D, McCollum C N, Lip G Y. Relationship between plasma markers of endothelial cell integrity and the Framingham cardiovascular disease risk‐factor scores in apparently healthy individuals. Blood Coagul Fibrinol 2002. 13513–518.518
21. Alli C, Mariotti G, Avanzini F. et al Studio sulla Pressione Arteriosa nell'Anziano(SPAA). Long‐term prognostic impact of repeated measurements over 1 year of pulse pressure and systolic blood pressure in the elderly. J Hum Hypertens 2005. 19355–363.363 [PubMed]
22. Beevers D G. Epidemiological, pathophysiological and clinical significance of systolic, diastolic and pulse pressure. J Hum Hypertens 2004. 18531–533.533 [PubMed]
23. Lip G Y H, Frison L, Grind M. The effect of hypertension on anticoagulated patients with atrial fibrillation. Eur Heart J 2006. (In press)
24. Michelson A D, Barnard M R, Hechtman H B. et al In vivo tracking of platelets: circulating degranulated platelets rapidly lose surface P‐selectin but continue to circulate and function. Proc Natl Acad Sci USA 1996. 9311877–11882.11882 [PubMed]
25. Blann A D, Lip G Y, Beevers D G. et al Soluble P‐selectin in atherosclerosis: a comparison with endothelial cell and platelet markers. Thromb Haemost 1997. 771077–1080.1080 [PubMed]
26. Fijnheer R, Frijns C J, Korteweg J. et al The origin of P‐selectin as a circulating plasma protein. Thromb Haemost 1997. 771081–1085.1085 [PubMed]
27. Connolly S, Pogue J, Hart R. et al ACTIVE Writing Group on behalf of the ACTIVE investigators; Clopidogrel plus aspirin versus oral anticoagulation for atrial fibrillation in the Atrial fibrillation Clopidogrel Trial with Irbesartan for prevention of Vascular Events (ACTIVE W): a randomised controlled trial. Lancet 2006. 3671903–1912.1912 [PubMed]
28. Chirinos J A, Castrellon A, Zambrano J P. et al Digoxin use is associated with increased platelet and endothelial cell activation in patients with nonvalvular atrial fibrillation. Heart Rhythm 2005. 2525–529.529 [PubMed]
29. Li‐Saw‐Hee F L, Blann A D, Lip G Y. Effects of fixed low‐dose warfarin, aspirin‐warfarin combination therapy, and dose‐adjusted warfarin on thrombogenesis in chronic atrial fibrillation. Stroke 2000. 31828–833.833 [PubMed]
30. Kamath S, Blann A D, Caine G J. et al Platelet P‐selectin levels in relation to plasma soluble P‐selectin and beta‐thromboglobulin levels in atrial fibrillation. Stroke 2002. 331237–1242.1242 [PubMed]
31. Li‐Saw‐Hee F L, Blann A D, Lip G Y. Effect of degree of blood pressure on the hypercoagulable state in chronic atrial fibrillation. Am J Cardiol 2000. 86795–7, A9.7, A9 [PubMed]
32. Boos C J, Lip G Y. Is hypertension an inflammatory process? Curr Pharm Des 2006. 121623–1635.1635 [PubMed]
33. Nadar S K, Tayebjee M H, Messerli F. et al Target organ damage in hypertension: pathophysiology and implications for drug therapy. Curr Pharm Des 2006. 121581–1592.1592 [PubMed]

Articles from Heart are provided here courtesy of BMJ Publishing Group