Känel, Roland von MD*†; Kudielka, Brigitte M PhD‡; Helfricht, Susanne PhD*; Metzenthin, Petra PhD§; Preckel, Daniel PhD*; Haeberli, André PhD¶; Cung, Trinh BS¶; Fischer, Joachim E MD, MSc∥
Acute mental stress and intense emotions particularly as relates to negative affect are now well-recognized triggers of acute coronary events.1,2 A prothrombotic milieu evolving during the initiation and development of a coronary thrombosis contributes to thrombus growth and thereby severity of myocardial damage.3,4 Several hemostatic factors reflecting a prothrombotic state have been found to be increased in the acute phase of a coronary event. Amongst these were von Willebrand factor antigen (VWF:Ag), fibrinogen, coagulation factor VII (FVII) and XII (FXII), and the coagulation activation marker fibrin D-dimer.5-8 Abundant evidence suggests that acute psychosocial stress elicits a hypercoagulable state9,10 that could add to the hypercoagulability due to the exposure of prothrombotic plaque material to the blood stream.11,12 Means targeting at reducing the prothrombotic state during acute coronary syndromes triggered by mental stress and emotions could thus potentially constrain cardiovascular threat acutely and in prognostic terms.
Aspirin and beta-blockers are part of the standard drug regimen of patients with coronary artery disease (CAD). Metaanalyses show that both of these drugs reduce recurrent CAD events and mortality by 20% to 30% in secondary prevention.13,14 The stress-induced prothrombotic state is partially mediated by catecholamines.15 In experimental human studies, beta-adrenergic agents were shown to activate the coagulation and fibrinolysis system and platelets.15,16 Therefore, beta-blocking agents and platelet-inhibiting drugs may be assumed to have a potential of mitigating the acute prothrombotic stress response.
We investigated the responsiveness of plasma levels of VWF:Ag, fibrinogen, FVII:C, FXII:C, and D-dimer to short-term psychosocial stress in apparently healthy volunteers who were randomized, double-blind, to aspirin, propranolol, aspirin and propranolol combined, or placebo medication. All hemostatic measures investigated in the present study have been shown to increase to acute stress in a male population similar in age to the sample investigated here17 and also in other samples of healthy subjects and patients with cardiovascular diseases.18-20 We previously showed in healthy subjects that stress-induced increase in these hemostatic factors had returned to pre-stress values at 45 minutes post-stress.17 We therefore did not extend our measurements beyond 45 minutes of recovery from stress.
We hypothesized that aspirin and propranolol would affect the change in plasma levels of hemostatic factors during stress and recovery differently from placebo. Because demographic factors, classic cardiovascular risk factors, and health habits have repeatedly shown to affect hemostasis,21-24 we controlled our analysis for gender, age, body mass index (BMI), screening blood pressure (BP), smoking status, sleep quality, alcohol consumption, and physical exercise. We postulated that our study medication would affect the stress response in hemostatic factors independent of these covariates of hemostatic function.
Study Participants and Recruitment
As part of a larger survey on Work & Health, we invited all 1802 permanent nonfaculty employees of the Federal Institute of Technology, Zürich, Switzerland older than 35 years of age to complete a structured health questionnaire inquiring about demographic variables, medical history, and health habits. A total of 355 employees returned the questionnaire of which 255 also consented for a routine laboratory work-up and participation in a study on the effects of psychosocial stress and medication on blood coagulation. Four different ethic committees approved the protocol of the medication study: Swiss Federal Institute of Technology, Zürich; Department of Internal Medicine, University of Zürich; State of Zürich; and Switzerland's regulatory agency for therapeutic products (Swissmedic).
Questionnaire data and laboratory charts were reviewed by a study physician who also took a history. On the basis of this information, candidates were eligible for the study if they appeared healthy and if they did not take any medication. We specifically excluded subjects with the following conditions: any hematological, pulmonary, liver, renal, gastrointestinal, heart, cerebrovascular, or psychiatric diseases; any history of thromboembolism, any current major or minor infection, or any trauma or surgery in the previous 6 months. We also excluded subjects who reported aspirin intolerance, known allergies to the study medication, and previous gastrointestinal bleedings.
Eligibility criteria were met by 76 subjects. Technical problems during the stress test (eg, delay in blood drawing) yielded unreliable data from 3 subjects leaving a sample of 73 subjects. Because of occasional assay problems and technical difficulties during the experiment (eg, clotted lines), hemostasis data were incomplete in 17 subjects. More precisely, plasma levels of VWF:Ag, fibrinogen, FVII:C, FXII:C, and D-dimer were missing 3, 3, 4, 3, and 4 times, respectively, immediately pre-stress, 11, 8, 8, 8, and 9 times, respectively, immediately post-stress, and 4 times each 45 minutes post-stress. We therefore report on 56 subjects who had a complete data set in terms of all 5 hemostatic measures at all time points assessed (ie, immediately pre-stress, immediately post-stress, and 45 minutes post-stress).
As part of a previous study protocol,25 we recruited 7 apparently healthy men similar to the subjects in the present study in terms of age (44 ± 4 years), BMI (26 ± 5 kg/m2), systolic BP (122 ± 8 mm Hg), diastolic BP (80 ± 6 mm Hg), and days of alcohol consumption per week (3.6 ± 2.1). Controls did not have experience with the TSST. In the nonstressed control group, blood was collected in the morning at identical time points and identically processed as in the stressed subjects. If levels of hemostatic factors do not significantly change in the control group, this would mean that circadian effects and blood processing method could be excluded as reasons for observed changes in hemostatic measures in the stressed group.
Demographic, Metabolic, and Life Style Factors
We extracted the data about gender, age, smoking history (current versus noncurrent smokers), alcohol consumption (average number of days/week subjects drank alcohol in the previous year), physical exercise (average hours/week subjects exercised in the previous year), and sleep quality (score ranging between 0 = very good sleep and 20 = very poor sleep) from the health questionnaire. With an interval of 3 minutes, 2 seated screening systolic and diastolic BP measurements were obtained using a sphygmomanometer. The mean arterial BP (MAP) was computed to be used in statistical analysis. The BMI was also calculated by dividing the weight in kilograms by the square of height in meters (kg/m2).
All capsules were fabricated in equal sizes and colors following standard procedures by the pharmacy of the University Hospital Zürich and delivered in numbered envelopes. Verum medication was 100 mg of acidum acetylsalicylicum (Aspirin cardio 100, Bayer Inc., Zürich, Switzerland) and 80 mg of propranololi hydrochloricum (Inderal LA 80, AstraZeneca Inc., Zug, Switzerland). Placebo pills contained lactose monohydrate (Hänseler AG, Herisau, Switzerland).
The study applied a randomized, double-blind, placebo-controlled, block design. The pills wrapped in envelopes were handed out to subjects who at this time were instructed to orally take 1 pill each day in the morning. Pills either contained placebo plus aspirin (n = 14), placebo plus propranolol (n = 15), aspirin plus propranolol (n = 16), or placebo plus placebo (n = 11). Two subjects did not consent for verum medication. They were not excluded but instead were given unblinded placebo/placebo pills to not critically reduce the cell size of the placebo group below ten subjects. Participants refrained from taking aspirin and any other nonsteroidal antiinflammatory drug within 10 days before the beginning of the study.
We used the standardized Trier-Social-Stress-Test (TSST) that combines a 3-minute preparation phase, a 5-minute free speech phase (mock job interview), and a 5-minute mental arithmetic task in front of an audience.26 The TSST evokes robust physiological stress responses,27 including in the hemostatic measures investigated in the present study.17
Participants were instructed not to have exercised and eaten on the morning of the testing day on which they took the last capsule. Testing started at 7:30 am. Either 2 or 3 subjects were scheduled to arrive at the laboratory with an interval of 30 to 45 minutes. About 10 minutes later, we equipped subjects with an indwelling venous cubital catheter and served a standardized light breakfast without caffeine. Thereafter, they remained seated and were informed about the nature of the stress protocol approximately one hour after arrival at the laboratory. After stress termination, subjects sat in a quiet room for a period of 45 minutes during which they watched a volume of landscape photography and travel magazines to prevent emotional arousal.
Blood samples for the hemostatic measures were obtained, and BP was measured using a sphygmomanometer immediately before the preparation phase (ie, pre-stress), immediately after the mental arithmetic task (ie, post-stress), and 45 minutes post-stress. Saliva samples for the measurement of saliva cortisol were collected immediately before stress, 15 minutes post-stress (ie, when the peak cortisol response is expected), and 45 minutes post-stress. For technical reasons, data for systolic and diastolic BP were incomplete in 5 subjects.
We added 4.5 mL of whole blood to 0.5 mL of CTAD-PPACK anticoagulant [stock solution containing 25 mL of citrate-theophylline-adenosine-dipyridamole (Becton Dickinson Biosciences, Allschwil, Switzerland) plus 5 mg of phenylalanyl-prolyl-arginine-chloromethylketone (Calbiochem-Novabiochem Inc., Läufelfingen, Switzerland)]. This anticoagulant prevents artificial coagulation activation by specific inhibition of thrombin.28 Samples were immediately centrifuged for 10 minutes at 3000 × g at room temperature. Plasma aliquots were frozen in polypropylene test tubes at -80°C until further analyses. Fibrinogen levels were quantified as per a modified Clauss method.29 FVII:C and FXII:C were determined by standard coagulometric methods using factor-deficient standard human plasma and reagents (Dade Behring, Liederbach, Germany). VWF:Ag was measured by a turbidimetric method (Dade Behring),30 and D-dimer was determined using a commercially available enzyme-linked immunosorbent assay (Asserachrom Stago, Asnières, France). All hemostatic measures were determined in duplicates, and all inter- and intra-assay coefficients of variation were less than 10%.
For cortisol measures, saliva was collected in purpose-designed tubes (IBL, Hamburg, Germany). Samples were frozen at -20°C. Thawed samples were centrifuged for 5 minutes at 3000 × g. Free cortisol levels were measured by a commercial luminescence immunoassay kit (IBL, Hamburg, Germany).31 Inter- and intra-assay coefficients of variation were less than 10%.
We used SPSS 13.0 for Windows to analyze the data. The significance level was set at P ≤ 0.05, and all testing was 2-tailed. Data are presented as means ± SD or means ± SEM. D-dimer levels were logarithmically transformed to obtain a normal data distribution. To test for differences in continuous and categorical data between groups, we applied 1-way analysis of variance (ANOVA) and covariance (ANCOVA) and Pearson chi-square test, respectively.
Repeated measures ANOVA was used to investigate the stress-induced changes in BP and cortisol with time points as the within-subject factor and medication groups as the between-subject factor. Repeated measures analysis of covariance (ANCOVA) was used to test whether changes over time in levels of hemostatic factors would be different between the 4 medication groups when controlling for covariates. We decided a priori to control for 6 covariates, namely gender, age, BMI, screening MAP, smoking status, and sleep quality to prevent model overfitting by inclusion of too many covariates given the sample size of 56 subjects.32 Only in a complementary analysis, we additionally controlled for alcohol consumption and physical exercise. To account for violations of the sphericity assumption, we applied the Huynh-Feldt correction for the degrees of freedom. Effect sizes are expressed by partial eta squared (ηp2). Post hoc comparisons used univariate ANCOVA and applied Fisher's Least Significant Difference.
Seventy-one percent of the 56 subjects were men. The mean age of the sample was 46.4 ± 7.7 years (range, 33 to 59 years), and the mean BMI was 25.2 ± 3.5 kg/m2. The mean screening SBP was 125 ± 16 mm Hg, and the mean DBP was 82 ± 9 mm Hg. Only 14% of subjects reported that they currently smoked, and only 14% reported that they did not weekly exercise. Together with moderate alcohol consumption on an average of 2.8 ± 2.0 days per week and a mean sleep quality score of 4.6 ± 3.6 indicating fairly good sleep, these data suggest that our subjects were reasonably healthy. Table 1 shows that demographic, metabolic, and lifestyle factors were not significantly different among medication groups.
Effect of Medication on Biological Measures Immediately Pre-Stress
Table 2 shows the significant differences in the pre-stress SBP level between medication groups. Post hoc comparisons revealed that subjects in the propranolol/placebo group had significantly lower SBP than those in the aspirin/propranolol group (P = 0.050), in the aspirin/placebo group (P = 0.002), and in the placebo/placebo group (P = 0.001). In contrast, the medication groups did not significantly differ in pre-stress levels of DBP, cortisol, and hemostasis factors.
Effect of Medication on Stress Responses in Biological Factors
Blood Pressure and Cortisol
In all subjects, the stress provoked a significant change over time in SBP (F2,94 = 23.2, P < 0.001), DBP (F2,94 = 13.1, P < 0.001), and cortisol (F2,85 = 28.2, P < 0.001) levels, thereby verifying that the TSST provoked a profound biological stress response. As previously reported, Figure 1 demonstrates that the stress-provoked changes in SBP (panel A) and in DBP (panel B),33 as well as in cortisol levels (panel C) (Kudielka et al, submitted) were not significantly different between medication groups.
In repeated measures ANCOVA, we adjusted the magnitude of change in hemostatic parameters for gender, age, BMI, screening MAP, smoking status, and sleep quality. Panel A of Figure 2 shows the significant stress-by-medication group interaction for the magnitude of change in VWF:Ag levels from immediately pre-stress to 45 minutes post-stress (F6,92 = 2.39, P = 0.035; ηp2 = 0.135). Post hoc analysis revealed a difference in VWF:Ag levels between immediately pre-stress and 45 minutes post-stress among the placebo/placebo group and the aspirin/propranolol group (P = 0.004), the aspirin/placebo group (P = 0.019), and the propranolol/placebo group (P = 0.010). The 3 groups on verum medication did not differ in their VWF:Ag level response (P > 0.47). More precisely, whereas VWF:Ag levels increased from pre-stress to 45 minutes after stress in the placebo/placebo group, all other groups experienced a decrease in their VWF:Ag levels during this time interval. In addition, the difference in VWF:Ag levels between immediately post-stress and 45 minutes post-stress was different between the aspirin/propranolol group and the placebo/placebo group (P = 0.012). During this time interval, VWF:Ag levels decreased in the aspirin/propranolol group and increased in the placebo/placebo group.
Figure 2 further shows that stress responses in fibrinogen (panel B; ηp2 = 0.028), FVII:C (panel C; ηp2 = 0.087), FXII:C (panel D; ηp2 = 0.013), and D-dimer (panel E; ηp2 = 0.020) levels were not significantly different between medication groups adjusting for covariates. In all subjects, there was, however, a main effect for stress on fibrinogen (F2,92 = 2.82, P < 0.07; ηp2 = 0.058) and FXII:C (F2,92 = 2.94, P < 0.06; ηp2 = 0.060) responses over time, both reaching borderline significance. In post hoc analysis, fibrinogen was higher immediately post-stress than 45 minutes post-stress (P = 0.023). FXII:C was higher immediately post-stress than immediately pre-stress (P = 0.003) and 45 minutes post-stress (P < 0.001).
The healthy middle-aged men in the nonstressed control group did not experience a significant change in any hemostatic measure during the test interval (data not shown in detail). This suggests that changes in hemostatic measures in the medication groups were evoked by the TSST and not by circadian changes or blood processing.
We additionally controlled the above repeated measures ANCOVA for alcohol consumption and physical exercise to explore whether these health habits affect the response to stress in any hemostatic measure. The significance of the interaction between stress and medication groups for the VWF:Ag stress response was maintained (F6,88 = 2.23, P = 0.047; ηp2 = 0.132). Stress did not significantly interact with medication groups in determining responses in fibrinogen, FVII:C, FXII:C, and D-dimer levels. Notably, this analysis considered 8 covariates, so that the models could be unstable due to overfitting.32
This study investigated whether 2 cardioprotective drugs might attenuate the prothrombotic response to acute psychosocial stress in middle-aged apparently healthy individuals controlling for previous correlates of hemostatic function. We found that subjects on placebo medication had a significant increase in plasma VWF:Ag levels from immediately before stress to 45 minutes after stress relative to subjects who had single aspirin, single propranolol, or the combination of the these 2 drugs. Furthermore, the increase in VWF:Ag levels during the recovery interval from immediately post-stress to 45 minutes afterwards was also significant in the placebo group relative to the group who received aspirin in combination with propranolol. In contrast, single aspirin, single propranolol, and aspirin combined with propranolol did not affect the response in plasma levels of fibrinogen, FVII:C, FXII:C, and D-dimer to acute stress differently from placebo medication. More specifically, even though fibrinogen and FXII:C levels changed during stress or the recovery interval, their changing seemed unaffected by our study medication.
The mitigating effect of aspirin and propranolol on the VWF:Ag response to stress has not previously been shown. However, there is a parallel line of research demonstrating that infusion of beta-mimetic agents results in regulated secretion of large and hemostatically active VWF molecules from vascular endothelial storage sites (ie, from the Golgi apparatus) into the circulation.15 Importantly, this effect occurs instantly, is dose-dependent, and seems relatively stronger in patients with atherosclerotic diseases than in healthy controls.15,16 Moreover, beta-adrenergic blockade prevented epinephrine-induced release of VWF from cultured human umbilical venous endothelial cells34,35 and acute increase in the plasma VWF concentration to infused epinephrine in healthy subjects.36 The TSST regularly achieves sympathomedullary activation as evidenced by a surge in plasma epinephrine levels,37 although nonselective beta-blockers (including propranolol) did not previously attenuate epinephrine increase in response to acute mental stress.38,39 We therefore propose that blunting of the acute VWF:Ag increase to stress in our subjects receiving propranolol relative to those on placebo was primarily a consequence of blocked vascular endothelial beta receptors inhibiting an epinephrine-triggered release of VWF molecules from the endothelium. However, we acknowledge that our data were unable to prove such a mechanism because we did not assess catecholamines and adrenergic receptor sensitivity in the present study. Also, propranolol did not affect the stress-induced increases in blood pressure and cortisol. Therefore, it could be argued whether the propranolol dose was sufficient to block VWF:Ag release. A randomized, double-blind crossover design investigating the stress response of VWF:Ag before and after administration of the study medication in the same patients could have helped resolving this dosage issue.
The role of aspirin in preventing a stress-induced increase in VWF:Ag levels is less clear. However, 5-day treatment with 100 mg/d aspirin suffices to evidently inhibit platelets.40 Several studies show that platelets are activated by epinephrine in vivo.15 Activated platelets release VWF from their storage granules and can even interact with intact endothelial cells via adhesion processes, thereby activating the endothelium to release VWF.41 It therefore could be that aspirin treatment mitigated epinephrine-triggered platelet activation during stress, thereby resulting in attenuated recruitment of VWF molecules from platelets and the endothelium.
The mechanisms by which acute psychosocial stress elicited a significant increase in fibrinogen, FVII:C, and FXII:C levels in plasma in previous studies have not been explored.9,15 Some authors propose that stress-hemoconcentration leads to a passive increase in circulating hemostatic factors.10 BP increase during stress determines part of stress-hemoconcentration.42 Because stress-induced BP changes were not different between the medication groups, we feel that stress-hemoconcentration is an unlikely explanation for the differences seen in VWF:Ag levels between our groups. D-dimer is a coagulation activation marker and as such a sensitive measure of a thrombogenic state.43 The molar concentration of D-dimer in plasma is regulated by several processes, amongst them are hepatic clearance and activity of the fibrinolytic system,44 both of which seem responsive to beta-adrenergic mechanisms.15,16 Greater levels of norepinephrine surge,45 cognitive stress appraisal,45 and anxiety19 have also been associated with higher plasma D-dimer concentration with acute stress. However, parsimoniously interpreted, our data do not suggest that beta-adrenergic mechanisms (including catecholamine spillover with stress) and aspirin by means of inhibiting platelet activity critically affected the stress responses in fibrinogen, FVII:C, and FXII:C. Such a notion concurs with previous studies in which aspirin insufficiently prevented platelet activation to norepinephrine infusion in healthy subjects46 and to physical exercise in patients with CAD.47 Moreover, the fact that aspirin and propranolol both mitigated the stress-induced increase in VWF:Ag levels but did not affect the stress response in fibrinogen, FVII:C, and FXII:C levels might not have sufficed to result in differences in overall fibrin formation (ie, D-dimer levels) between medication groups.
Our findings may have clinical implications in that a procoagulant milieu in response to acute psychosocial stress and intense negative affect is thought to contribute to the propagation of coronary thrombosis after plaque rupture.9-12 Many patients with CAD take aspirin and a beta-blocking agent, though not necessarily a nonselective beta-blocker. If aspirin and nonselective beta-blockade do not weaken the increase in several prothrombotic factors with acute stress, they might have only limited use for reducing the coronary risk with acute stress and emotions. However, the coronary benefit of these drugs is not to be considered as null. For instance, the risk of myocardial infarction onset after an intense bout of anger was lower in patients who were regular aspirin users than in those who did not use aspirin.48 Our data may suggest a relatively blunted VWF:Ag increase with the stress of intense anger experienced by those patients with CAD who were on aspirin.
The possibility that our study had limited statistical power to detect more subtle differences in the hemostatic response between medication groups should be bared in mind. We point out that our subjects were apparently healthy and comparably younger than the average patient with a heart disease, thereby not allowing us to draw a firm conclusion from our data to populations of patients with CAD. Moreover, we are unable to preclude that other, particularly higher, dosages of aspirin and propranolol might evidently affect the acute stress response in plasma levels of fibrinogen, FVII:C, FXII:X, and D-dimer.
Aspirin and nonselective beta-blockade may attenuate the response in plasma VWF:Ag levels to acute psychosocial stress. However, plasma levels of fibrinogen, FVII:C, FXII:C, and D-dimer did not differ in their response to stress between the verum and placebo medication groups. This may suggest that aspirin and propranolol only partly mitigate the acute prothrombotic stress response in healthy subjects. The clinical implications of our findings for patients with CAD are to be elucidated.
This work was carried out while several authors were affiliated with the former Institute of Behavioral Sciences at the Swiss Federal Institute of Technology, Zürich. The institute was closed on the death of its chairman, Professor Karl Frey. We feel indebted to his support during the conduct of this study.
1. Bhattacharyya MR, Steptoe A. Emotional triggers of acute coronary syndromes: strength of evidence, biological processes, and clinical implications. Prog Cardiovasc Dis. 2007;49:353-365.
2. Servoss SJ, Januzzi JL, Muller JE. Triggers of acute coronary syndromes. Prog Cardiovasc Dis. 2002;44:369-380.
3. Virmani R, Kolodgie FD, Burke AP, et al. Lessons from sudden coronary death: a comprehensive morphological classification scheme for atherosclerotic lesions. Arterioscler Thromb Vasc Biol. 2000;20:1262-1275.
4. Turpie AG, Antman EM. Low-molecular-weight heparins in the treatment of acute coronary syndromes. Arch Intern Med. 2001;161:1484-1490.
5. Heper G, Bayraktaroglu M. The importance of von Willebrand factor level and heart rate changes in acute coronary syndromes: a comparison with chronic ischemic conditions. Angiology. 2003;54:287-299.
6. Bayes-Genis A, Mateo J, Santalo M, et al. D-Dimer is an early diagnostic marker of coronary ischemia in patients with chest pain. Am Heart J. 2000;140:379-384.
7. Campo G, Valgimigli M, Ferraresi P, et al. Tissue factor and coagulation factor VII levels during acute myocardial infarction: association with genotype and adverse events. Arterioscler Thromb Vasc Biol. 2006;26:2800-2806.
8. Grundt H, Nilsen DW, Hetland O, et al. Activated factor 12 (FXIIa) predicts recurrent coronary events after an acute myocardial infarction. Am Heart J. 2004;147:260-266.
9. von Kanel R, Mills PJ, Fainman C, et al. Effects of psychological stress and psychiatric disorders on blood coagulation and fibrinolysis: a biobehavioral pathway to coronary artery disease? Psychosom Med. 2001;63:531-544.
10. Thrall G, Lane D, Carroll D, et al. Systematic review of the effects of acute psychological stress and physical activity on haemorheology, coagulation, fibrinolysis and platelet reactivity: Implications for the pathogenesis of acute coronary syndromes. Thromb Res. 2007;120:819-847.
11. Gidron Y, Gilutz H, Berger R, et al. Molecular and cellular interface between behavior and acute coronary syndromes. Cardiovasc Res. 2002;56:15-21.
12. Strike PC, Magid K, Whitehead DL, et al. Pathophysiological processes underlying emotional triggering of acute cardiac events. Proc Natl Acad Sci USA. 2006;103:4322-4327.
13. Collaborative overview of randomised trials of antiplatelet therapy-I: Prevention of death, myocardial infarction, and stroke by prolonged antiplatelet therapy in various categories of patients. Antiplatelet Trialists' Collaboration. BMJ. 1994;308:81-106.
14. Everly MJ, Heaton PC, Cluxton RJ Jr. Beta-blocker underuse in secondary prevention of myocardial infarction. Ann Pharmacother. 2004;38:286-293.
15. von Kanel R, Dimsdale JE. Effects of sympathetic activation by adrenergic infusions on hemostasis in vivo. Eur J Haematol. 2000;65:357-369.
16. von Kanel R, Dimsdale JE, Adler KA, et al. Effects of non-specific beta-adrenergic stimulation and blockade on blood coagulation in hypertension. J Appl Physiol. 2003;94:1455-1459.
17. von Kanel R, Preckel D, Zgraggen L, et al. The effect of natural habituation on coagulation responses to acute mental stress and recovery in men. Thromb Haemost. 2004;92:1327-1335.
18. Jern C, Eriksson E, Tengborn L, et al. Changes of plasma coagulation and fibrinolysis in response to mental stress. Thromb Haemost. 1989;62:767-771.
19. von Kanel R, Dimsdale JE, Adler KA, et al. Effects of depressive symptoms and anxiety on hemostatic responses to acute mental stress and recovery in the elderly. Psychiatry Res. 2004;126:253-264.
20. Wirtz PH, Ehlert U, Emini L, et al. Acute procoagulant stress reactivity and recovery in apparently healthy men with systolic anddiastolic hypertension. J Psychosom Res. 2007;63:51-58.
21. Folsom AR. Hemostatic risk factors for atherothrombotic disease: an epidemiologic view. Thromb Haemost. 2001;86:366-373.
22. Kannel WB. Overview of hemostatic factors involved in atherosclerotic cardiovascular disease. Lipids. 2005;40:1215-1220.
23. Lee KW. Lip GY. Effects of lifestyle on hemostasis, fibrinolysis, and platelet reactivity: a systematic review. Arch Intern Med. 2003;163:2368-2392.
24. von Kanel R, Loredo JS, Ancoli-Israel S, et al. Association between polysomnographic measures of disrupted sleep and prothrombotic factors. Chest. 2007;131:733-739.
25. von Kanel R, Kudielka BM, Preckel D, et al. Delayed response and lack of habituation in plasma interleukin-6 to acute mental stress in men. Brain Behav Immun. 2006;20:40-48.
26. Kudielka BM, Wüst S, Kirschbaum C, et al. Trier Social Stress Test. In: G Fink (Ed.), Encyclopedia of Stress, 2nd revised edition. Oxford: Academic Press, 2007.
27. Dickerson SS, Kemeny ME. Acute stressors and cortisol responses: a theoretical integration and synthesis of laboratory research. Psychol Bull. 2004;130:355-391.
28. Herren T, Stricker H, Haeberli A, et al. Fibrin formation and degradation in patients with arteriosclerotic disease. Circulation. 1994;90:2679-2686.
29. Clauss A. Gerinnungsphysiologische Schnellmethode zur Bestimmung des Fibrinogens. Acta Haematol. 1957;17:237-246.
30. Sukhu K, Martin PG, Cross L, et al. Evaluation of the von Willebrand factor antigen (vWF-Ag) assay using an immuno-turbidimetric method (STA Liatest vWF) automated on the MDA 180 coagulometer. Clin Lab Haematol. 2000;22:29-32.
31. Westermann J, Demir A, Herbst V. Determination of cortisol in saliva and serum by a luminescence-enhanced enzyme immunoassay. Clin Lab. 2004;50:11-24.
32. Babyak MA. What you see may not be what you get: a brief, nontechnical introduction to overfitting in regression-type models. Psychosom Med. 2004;66:411-421.
33. von Kanel R, Kudielka BM, Metzenthin P, et al. Aspirin, but not propranolol, attenuates the acute stress-induced increase in circulating levels of interleukin-6: A randomized, double-blind, placebo-controlled study. Brain Behav Immun. (in press).
34. Wall RT, Counts RB, Harker LA, et al. Binding and release of factor VIII/von Willebrand's factor by human endothelial cells. Br J Haematol. 1980;46:287-298.
35. Vischer UM, Wollheim CB. Epinephrine induces von Willebrand factor release from cultured endothelial cells: involvement of cyclic AMP-dependent signalling in exocytosis. Thromb Haemost. 1997;77:1182-1188.
36. Larsson PT, Wallen NH, Martinsson A, et al. Significance of platelet beta-adrenoceptors for platelet responses in vivo and in vitro. Thromb Haemost. 1992;68:687-693.
37. Schommer NC, Hellhammer DH, Kirschbaum C. Dissociation between reactivity of the hypothalamus-pituitary-adrenal axis and the sympathetic-adrenal-medullary system to repeated psychosocial stress. Psychosom Med. 2003;65:450-460.
38. Paran E, Neumann L, Cristal N. Effects of mental and physical stress on plasma catecholamine levels before and after beta-adrenoceptor blocker treatment. Eur J Clin Pharmacol. 1992;43:11-15.
39. Freyschuss U, Hjemdahl P, Juhlin-Dannfelt A, et al. Cardiovascular and sympathoadrenal responses to mental stress: influence of beta-blockade. Am J Physiol. 1988;255:H1443-H1451.
40. Schror K. Aspirin and platelets: the antiplatelet action of aspirin and its role in thrombosis treatment and prophylaxis. Semin Thromb Hemost. 1997;23:349-356.
41. Chen J, Lopez JA. Interactions of platelets with subendothelium and endothelium. Microcirculation. 2005;12:235-246.
42. Allen MT, Patterson SM. Hemoconcentration and stress: a review of physiological mechanisms and relevance for cardiovascular disease risk. Biol Psychol. 1995;41:1-27.
43. Lip GY, Lowe GD. Fibrin D-dimer: A useful marker for thrombogenesis? Clin Sci. 1995;89:205-214.
44. Chandler WL, Velan T. Plasmin generation and D-dimer formation during cardiopulmonary bypass. Blood Coagul Fibrinolysis. 2004;15:583-591.
45. Wirtz PH, Ehlert U, Emini L, et al. Anticipatory cognitive stress appraisal and the acute procoagulant stress response in men. Psychosom Med. 2006;68:851-858.
46. Larsson PT, Wallen NH, Hjemdahl P. Norepinephrine-induced human platelet activation in vivo is only partly counteracted by aspirin. Circulation. 1994;89:1951-1957.
47. Wallen NH, Held C, Rehnqvist N, et al. Effects of mental and physical stress on platelet function in patients with stable angina pectoris and healthy controls. Eur Heart J. 1997;18:807-815.
48. Mittleman MA, Maclure M, Sherwood JB, et al. Triggering of acute myocardial infarction onset by episodes of anger. Determinants of Myocardial Infarction Onset Study Investigators. Circulation. 1995;92:1720-1725.
© 2008 Lippincott Williams & Wilkins, Inc.