| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
ORIGINAL ARTICLES |
From the Departments of Psychiatry and Neuropsychology (R.v.D.), Hematology (K.H., C.v.Z.), Medical, Clinical, and Experimental Psychology (A.A.), Maastricht University, Maastricht, The Netherlands; and the Department of Medical and Clinical Psychology (W.J.K.), Uniformed Services University of the Health Sciences, Bethesda, MD.
Address reprints requests to: R. van Diest, Department of Psychiatry and Neuropsychology, Maastricht University, Box 616, 6200 MD Maastricht, The Netherlands. Email: rob.vandiest{at}pn.unimaas.nl
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTSDISCUSSIONACKNOWLEDGMENTSREFERENCES |
|---|
METHODS: Blood was drawn from 29 VE and 30 control males, all healthy and nonsmokers, to assess hemostatic measures at 7:00 AM and 6:00 PM.
RESULTS: All measures of fibrinolysis were in their normal range and showed significant diurnal variations. These variations were more pronounced in VE as all fibrinolytic measures were significantly higher in VE at 7:00 AM and similar to those of controls at 6:00 PM, thus supporting our hypothesis with respect to fibrinolysis. Diurnal decreases in tPA and tPA-PAI ranged from 1.5 (VE) to 1.3 (controls), whereas the diurnal decrease in PAI-1 was more than fourfold in VE and 2.7-fold in controls. This suggests a decreased fibrinolytic capacity in VE during the early morning. All coagulation measures were in their normal range, and prothrombin fragment 1+2 (F1+2), thrombin-antithrombin complexes, and activated factor VII showed significant diurnal variations. These variations were similar in VE and control individuals, thus not supporting our hypothesis with respect to coagulation. Finally, F1+2 and fibrinogen were both significantly higher throughout the day in VE.
CONCLUSIONS: VE is associated with decreased early morning fibrinolysis and increased fibrinogen levels throughout the day. These hemostatic changes may promote thrombus formation and provide a potential pathophysiological mechanism by which VE is related to MI and its circadian variation.
Key Words: stress, • coagulation, • fibrinolysis, • diurnal variations, • myocardial infarction.
Abbreviations: AF = alkaline phosphatase;; ALAT = alanin-aminotransferase;; ASAT = aspartic-aminotransferase;; BMI = body mass index;; CAD = coronary artery disease;; CV = intraassay coefficient of variation;; EDTA = ethylenediaminetetraacetic;; ELISA = enzyme-linked immunosorbent assay;; FNG = fibrinogen;; F1+2 = prothrombin fragment 1+2;; 
-GT = gamma-glutamyl transferase;; GSQ = Groningen sleep questionnaire;; LDH = lactate dehydrogenase;; Ldlevel = lower detection level;; MI = myocardial infarction;; MIVE = Maastricht interview vital exhaustion;; MQ = Maastricht questionnaire;; PAI-1:act = plasminogen activator inhibitor activity;; PSS = perceived stress scale;; RR = relative risk;; SCID = structured clinical interview for DSM-IV;; SEM = standard error of the mean;; TAT = thrombin-antithrombin complexes;; tPA:ag = tissue plasminogen activator antigen;; tPA-PAI:ag = tPA-PAI complexes’ antigen;; VE = vital exhaustion;; VII:a = activated factor VII;; VII:c, VIII:c = factor VII and VIII coagulant activity;; vWf:ag = von Willebrand factor antigen.
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONACKNOWLEDGMENTSREFERENCES |
|---|
Coagulation and fibrinolysis play an important role in the development of CAD (12) and may underlie the association of psychological distress with future cardiac events (13). Acute psychological stress, for instance, induces simultaneous increases in coagulability and fibrinolytic capacity (14), a finding that was not observed in VE. Instead, an increased activation of coagulation and a decreased fibrinolytic capacity have been associated with VE (15, 16). These hemostatic changes are likely to reflect a maladaptive response to chronic stress that enhances thrombus formation, thus promoting the risk of MI. Those studies, however, were confined to a limited set of measures to assess hemostasis, therefore providing only a partial representation of the complex processes involved in blood coagulation and fibrinolysis.
The high incidence of MI in the early morning is of additional importance (17) because waking up exhausted has been shown to be an independent risk indicator of MI (18). Although the morning increase in MI may be due to diurnal variations in coagulation and fibrinolysis (19), an interaction between VE and these diurnal variations has not yet been investigated. In the present study, the hypothesis is tested that VE is associated with more pronounced diurnal variations in hemostatic measures.
|
METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONACKNOWLEDGMENTSREFERENCES |
|---|
18) and 33 control subjects (MQ 
7) entered phase 2, in which the MIVE and the SCID were administered. The MIVE is a standardized interview of 23 questions, designed to assess vital exhaustion. Because the validity of the MIVE has been established in men only, no women were included in this study (21). The SCID is a standardized interview to assess psychopathology according to DSM criteria (22). Only the part assessing major depression was used to lower the burden of the study for the participants and because no other affective disorders are found to belong to the precursors of a coronary event. Subjects were classified as exhausted if a) they endorsed more than 7 of the 23 MIVE items and b) at least one of the elements of VE (fatigue, irritability, or general malaise) had increased in intensity over the past 18 months. On the basis of their responses during this second phase, 19 subjects were excluded because they did not meet MIVE criteria for VE. Three VE subjects were excluded because they also met DSM criteria for major depression or other current mood disorders. All controls met the selection criteria. Five selected subjects (VE = 2, controls = 3) could not participate in the main study on the scheduled dates. Enrollment was discontinued when a group of 30 exhausted subjects and 30 controls agreed to participate. One VE subject, however, was excluded post hoc due to an as yet undetected hypercholesterolemia (cholesterol = 9.5 mM/liter, normal range = 4.1–6.4; HDL = 0.6 mM/liter, normal = 0.6–1.9; LDL = 7.9, normal = 3.0–5.2). Thus, the final sample consisted of 29 VE and 30 control participants, all nonsmokers and in good health. The study protocol was approved by the Institutional Review Board, and all participants gave written informed consent.
During phase 2, additional assessments were made of daily alcohol and coffee consumption, weekly physical exercise (using a four-point scale from ‘no regular physical exercise’ to ‘daily physical exercise’), and habitual sleep quality over the past 3 months with the GSQ (6). This questionnaire consist of 14 yes/no items, with the higher the score (range 0–14), the worse the sleep quality. For validation purposes, the participants also appraised the level of stress experienced during the last month using the 10-item PSS (23).
Experimental Procedure
Each participant slept two consecutive nights in a sound-attenuated single room at the University Hospital. The first night was an adaptation night. Following a day of usual activities, participants returned to the hospital at 5:30 PM. They refrained from food, alcohol, and caffeinated beverages starting at 2:00 PM. Heart rate and blood pressure were measured after a bedrest of 15 minutes, and blood was withdrawn from an antecubital vein at 6:00 PM. A light dinner (3614 kJ) was served at 7:00 PM. After an overnight fast, the second venipuncture was performed at 7:00 AM with the participants in supine position before they engaged in physical activity. Because blood sampling with vacuum tubes may induce sampling artifacts (24), an open system with a 1.2-mm needle was used on both occasions, and a minimal stasis was applied to the upper arm.
Laboratory Techniques
Complete blood count (erythrocyte, thrombocyte, and leukocyte count, differential, hemoglobin, hematocrit) was assessed from EDTA blood collected at 6:00 PM and 7:00 AM (Gen-S automate, Beckman and Coulter, the Netherlands). Routine blood chemistry (AF, ALAT, ASAT, bilirubin, creatinine, -GT, LDH, lipids) was analyzed from serum collected at 7:00 AM (Synchron-CX automate, Beckman and Coulter). Glucose was measured from blood collected in potassium oxalate sodium fluoride tubes. Measures of coagulation were F1+2, TAT, VII:a, VII:c, VIII:c, vWf:ag, and FNG, and measures of fibrinolysis were tPA:ag, PAI-1:act, and tPA-PAI:ag. Blood for these measures was collected in plastic tubes containing a 1/10 volume of 0.129 M trisodium citrate. Tubes were immediately placed on melting ice and centrifuged within 10 minutes at 2000 g for five minutes and 6700 g for 10 minutes at 4°C. Aliquots of platelet-free plasma were snap frozen and stored at -70°C for further analysis.
F1+2 was measured with an ELISA (Dade Behring, Marburg, Germany; CV = 5–7.5% at the level of 0.2–5 nM/liter; ldlevel = 0.04 nM/liter, normal range = 0.4–1.1 nM/liter) and TAT with an ELISA using antibodies from Kordia-Biopool (Leiden, The Netherlands; CV = 4–6%; ldlevel = 0.5 µg/liter, normal range < 4 µg/liter). Factor VII:a was measured with a clotting assay of VII:a (Diagnostica Stago, Asnières-sur-Seine, France; CV = 4–6%; ldlevel = 5 mU/ml; normal range 25–65 mU/ml) and VII:c and VIII:c with a one-stage clotting assay using immunodepleted factor deficient reagents (Organon Teknika; Durham, NC; CV = 8%; ldlevel for VII:c = 5%, normal range = 60–140%; ldlevel for VIII:c = 1%, normal range = 50–180%). Data for VII:c and VIII:c are presented as percentages of values found with normal pool plasma. vWf:ag was measured with an LIA test (Diagnostica Stago, Mannheim, Germany; CV 4% at the level 70%; ldlevel = 1%, normal range = 50–200%) and clottable FNG according to a modified method of Clauss (CV < 5%; ldlevel = 0.1 g/liter, normal range = 1.7–4.0 g/liter) (25). tPA:ag and tPA-PAI:ag were measured with ELISAs (tPA:ag: Chromogenix, Mölndal, Sweden; CV = 7% at the level <10 ng/ml; ldlevel = 0.5 ng/ml, normal range = 1.7–6.5 ng/ml; tPA-PAI:ag: Kordia-Biopool, Leiden, The Netherlands; CV = 6.8% at the level of 2.5 ng/ml; ldlevel = 0.25 ng/ml, normal range = 0.6–6.7 ng/ml), and PAI-1:act was measured with a chromogenic assay (Kordia-Biopool; CV < 10% at the level <15 U/ml; ldlevel = 0 U/ml, normal range < 16 U/ml). The automate for VII:a, VII:c, and VIII:c was the ACL 300 (Instrumentation Laboratories, IJsselstein, The Netherlands) and for FNG and vWf:ag the STA (Diagnostica Stago, Mannheim, Germany). Laboratory personnel were blinded to the psychological status of the subjects, and all assays were analyzed in one batch to prevent interassay variation.
Statistical Analysis
Differences in diurnal variations between the two groups were analyzed using a mixed model analysis of variance with time (7:00 AM vs. 6:00 PM) as two-level within-subjects factor and group (VE vs. controls) as between-subjects factor. Significant group x time interactions were analyzed with post hoc t tests to examine whether significant differences between the two groups occurred at 7:00 AM or 6:00 PM. Triglycerides and (HDL and LDL) cholesterol were used as covariates to test whether group differences in coagulation and fibrinolysis could be accounted for by these lipids. Interrelations between these lipids and coagulation/fibrinolysis measures were examined with product-moment correlations. To explore a possible association between sleep problems and coagulation/fibrinolysis measures, t tests of independent samples were used. Absence of sleep problems was defined as a GSQ score 3 and presence of sleep problems as a GSQ score 4 (6). Natural logarithms were used to normalize skewed distributions as appropriate. Data are presented as mean ± SEM. Significance levels were based on two-tailed tests, with 
level set at .05.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONACKNOWLEDGMENTSREFERENCES |
|---|
|
|
Fibrinolytic measures were significantly interrelated (all r values .66). Significant correlations were also found between these measures and VII:c (all r values .48), between VII:c and VII:a (r = .54), and between VIII:c and vWf:ag (r = .86). Triglycerides were positively correlated with all fibrinolytic measures (all r values .57) and VII:c (r = .45), and HDL-cholesterol was negatively correlated with tPA-PAI:ag (r = -.44), PAI-1:act (r = -.36), and FNG (r = -.32).
More than 72% of the VE participants reported habitual sleep problems over the past 3 months, which is in sharp contrast with the 3% of controls that also reported to have habitual sleep problems (Table 1). Although this high multicollinearity prevents us from estimating the separate contributions of VE and sleep problems on coagulation/fibrinolysis measures, an exploratory analysis of the possible impact of disturbed sleep on hemostasis is warranted. That analysis showed that sleep problems are associated with higher levels of FNG (p = .04), PAI-1:act (p = .003), and tPA-PAI:ag (p = .03) at 7:00 AM, whereas no associations were found between habitual sleep problems and 6:00 PM samples of these same measures (all p values > .10).
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONACKNOWLEDGMENTSREFERENCES |
|---|
Compared with controls, however, VE individuals did demonstrate significantly higher levels of F1+2 and FNG at 7:00 AM and 6:00 PM. The elevation in F1+2, however, was not paralleled by a similar elevation in TAT, a discrepancy that may be due to differences in the clearance rates of F1+2 and TAT and to differences in the specificity of the detection methods. Furthermore, the considerable variability in TAT complexes formation, as reflected in the high SEMs of this measurement, hampers a direct association between F1+2 and TAT. Moreover, F1+2 levels in both groups were significantly lower at 7:00 AM, indicating that, if F1+2 is to represent thrombotic potential, this potential is actually decreased in the early morning. This implies that other factors, not studied here, may be involved in the circadian pattern of MI. Finally, the higher level of F1+2 in VE individuals remained well within the normal range. Although the increased amount of thrombin formed in VE individuals may promote the turnover of FNG into fibrin, thus increasing the risk of future MI (26), the relevance of F1+2 as a predictor of CAD has recently been challenged (27), suggesting that an increased prothrombin activation may be of minor importance in terms of the increased risk of CAD in exhausted individuals.
The higher level of tPA-PAI:ag at 7:00 AM indicates that the increased tPA in the early morning in VE is, for the most part, neutralized by its main inhibitor, PAI-1 (28). It is also important to note that the average diurnal decreases in tPA:ag and tPA-PAI:ag in VE individuals are about 1.5-fold, whereas their average diurnal decrease in PAI-1 activity is more than fourfold. In the control group, the average diurnal decreases in tPA:ag and tPA-PAI:ag are about 1.3-fold, but their average diurnal decrease in PAI-1 activity is about 2.7-fold. This differential decrease in PAI-1 activity (ie, higher levels of PAI-1 activity at 7:00 AM in VE individuals) suggests that VE is associated with a decreased fibrinolytic capacity during the early morning. Because increased fibrinogen levels throughout the day and decreased fibrinolytic capacity in the early morning are likely to promote thrombus formation, we suggest that these hemostatic changes provide a potential pathophysiological mechanism by which VE is related to MI and its circadian variation.
The physiological mechanisms underlying these hemostatic changes in VE remain to be elucidated. An increased adrenocortical responsiveness and a decreased fibrinolytic capacity in VE have been proposed to reflect the presence of the insulin resistance syndrome (ie, hypertension, obesity, dyslipidemia, and hyperinsulinemia) (29). However, despite the markers of decreased fibrinolysis among exhausted individuals in the present study, blood pressure, body mass index, and lipid levels were similar in VE vs. controls, congruent with previous observations in cardiac patients (10). Furthermore, it was recently shown that exhausted individuals were similar to controls in their adrenocortical responsiveness to psychological stress, whereas basal cortisol levels were significantly lower in VE (30). Low cortisol levels have been associated with elevated PAI-1 levels (31), which is of interest in future studies that examine the relation between coagulation, fibrinolysis, and activity of the hypothalamic-pituitary-adrenocortical axis in VE.
More than 72% of the VE participants reported habitual sleep problems over the past 3 months, which is in sharp contrast with the 3% of controls that also reported to have habitual sleep problems. Sleep problems have been implicated as a risk factor of first MI (5), and even a modest reduction in sleep duration enhances sympathoadrenal activity (32). Adrenergic stimulation may promote the production or release of tPA:ag (33) but does not significantly affect F1+2 (34). Our exploratory analysis showed that sleep problems are associated with higher levels of FNG, PAI-1:act, and tPA-PAI:ag at 7:00 AM, whereas no associations were found between sleep problems and 6:00 PM samples of these same measures. We did not assess sympathoadrenal activity, and future studies are therefore needed to investigate whether adrenergic stimulation provides a potential link between VE, disturbed sleep, and future MI.
Several lines of evidence indicate that major depression is a risk indicator for both the development of CAD and recurrent events after MI (35). Because major depression and VE have important symptoms in common, most notably the increased fatigue and irritability, it has been suggested that they are associated with the same underlying pathophysiological mechanisms with respect to future MI (36). We have previously demonstrated that VE individuals rarely reported any depressed mood during 3 weeks of daily monitoring with the Profile of Mood States, whereas they were significantly more tired than controls throughout the entire period of observation (37). Moreover, subjects suffering from current mood disorders were explicitly excluded from the present study. Our findings, therefore, do not necessarily apply to major depression, even more so because coagulation is probably not disordered in depressed individuals (38) and conflicting results exist with respect to fibrinolysis (13, 39).
In conclusion, our data show that VE is associated with increased fibrinogen levels throughout the day and decreased fibrinolytic capacity in the early morning. These hemostatic changes are likely to promote thrombus formation and provide a potential pathophysiological mechanism by which VE is related to MI and its circadian variation.
|
ACKNOWLEDGMENTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONACKNOWLEDGMENTSREFERENCES |
|---|
Received for publication March 22, 2001.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONACKNOWLEDGMENTSREFERENCES |
|---|
This article has been cited by other articles:
 |
|
 |
|
  B. T. Mausbach, R. von Kanel, K. Aschbacher, S. K. Roepke, J. E. Dimsdale, M. G. Ziegler, P. J. Mills, T. L. Patterson, S. Ancoli-Israel, and I. Grant Spousal Caregivers of Patients With Alzheimer's Disease Show Longitudinal Increases in Plasma Level of Tissue-Type Plasminogen Activator Antigen Psychosom Med, October 1, 2007; 69(8): 816 - 822. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. R. Rudnicka, A. Rumley, G. D.O. Lowe, and D. P. Strachan Diurnal, Seasonal, and Blood-Processing Patterns in Levels of Circulating Fibrinogen, Fibrin D-Dimer, C-Reactive Protein, Tissue Plasminogen Activator, and von Willebrand Factor in a 45-Year-Old Population Circulation, February 27, 2007; 115(8): 996 - 1003. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. van der Ven, R. van Diest, K. Hamulyak, M. Maes, C. Bruggeman, and A. Appels Herpes Viruses, Cytokines, and Altered Hemostasis in Vital Exhaustion Psychosom Med, March 1, 2003; 65(2): 194 - 200. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
) and controls (
). Data represent mean ± SEM. The coagulation measures F1+2, TAT, and VII:a and all fibrinolytic measures displayed significant diurnal variations (all p values < .01). Exhausted individuals, however, had significantly higher levels of tPA:ag, PAI-1:act, and tPA-PAI:ag at 7:00 AM (all p values < .04), whereas these levels were similar to those of controls at 6:00 PM (all p values > .10). This indicates a more pronounced diurnal variation in VE with respect to fibrinolytic measures. A similar, more pronounced diurnal variation in VE with respect to coagulation measures was absent (all p values > .10). Significantly higher levels of F1+2 and FNG were found in VE both at 7:00 AM and at 6:00 PM (p = .02), and a similar trend was found for VII:c (p = .07). No significant group differences were found in TAT, VII:a, VIII:c, and vWf:ag (all p values > .10).