Association of vital exhaustion and depressive symptoms with changes in fibrin D-dimer to acute psychosocial stress

Journal of Psychosomatic Research 67 (2009) 93 – 101

Association of vital exhaustion and depressive symptoms with changes in fibrin D-dimer to acute psychosocial stress Roland von Känel a , Silja Bellingrath b,c , Brigitte M. Kudielka b,c,⁎ a

Division of Psychosomatic Medicine, Inselspital, Bern University Hospital, and University of Bern, Switzerland b Jacobs Center on Lifelong Learning and Institutional Development, Jacobs University Bremen, Germany c Department of Theoretical and Clinical Psychobiology, Graduate School of Psychobiology, University of Trier, Germany Received 28 August 2008; received in revised form 18 November 2008; accepted 11 December 2008

Abstract Objective: Vital exhaustion and depression are psychosocial risk factors of coronary artery disease. A hypercoagulable state in response to acute psychosocial stress contributes to atherothrombotic events. We aimed to investigate the hypothesis that vital exhaustion and depression correlate with stress-induced changes in the hypercoagulability marker D-dimer. Methods: Thirty-eight healthy and nonsmoking school teachers (mean age 50±8 years, 55% women) completed the nine-item Maastricht Vital Exhaustion Questionnaire and the seven-item depression subscale of the Hospital Anxiety and Depression Scale. Within 1 week, subjects twice underwent the Trier Social Stress Test (i.e., preparation phase, mock job interview, and mental arithmetic that totaled 13 min). Plasma D-dimer levels were determined at five time points during the protocol. Results: Vital exhaustion (P=.022; η2=.080) and depressive symptoms (P=.011; η2=.090) were associated with stress-induced changes

in D-dimer levels over time controlling for sex and age. Elevated levels of vital exhaustion (r=−.46, P=.005) and of depression (r=−.51, P=.002) correlated with reduced D-dimer increase from pre-stress to immediately post-stress. Also, elevated vital exhaustion (r=.34, P=.044) and depression (r=.41, P=.013) were associated with increase (i.e., attenuated recovery) of D-dimer levels between 20 and 45 min post-stress. Controlling for stress hormone and blood pressure reactivity did not substantially alter these results. Conclusion: The findings suggest an attenuated immediate D-dimer stress response and delayed recovery of D-dimer levels post-stress with elevated vital exhaustion and depressive symptoms. In particular, the prolonged hypercoagulability after stress cessation might contribute to the atherothrombotic risk previously observed with vital exhaustion and depression, even at subclinical levels. © 2009 Elsevier Inc. All rights reserved.

Keywords: Blood coagulation; Cardiovascular disease; Depression; Psychological stress

Introduction Psychosocial factors such as vital exhaustion (VE) and depression are associated with an increased risk of coronary artery disease (CAD) [1–3]. Vital exhaustion has been conceptualized as a state of undue fatigue, loss of energy, increased irritability, and demoralization reflecting a breakdown of the adaptation to chronic stress [4]. The concept of ⁎ Corresponding author. Jacobs Center on Lifelong Learning and Institutional Development, Jacobs University Bremen, Campus Ring 1 / D28759 Bremen, Germany. Tel.: +49 421 200 4730; fax: +49 421 200 4793. E-mail address: [email protected] (B.M. Kudielka). 0022-3999/08/$ – see front matter © 2009 Elsevier Inc. All rights reserved. doi:10.1016/j.jpsychores.2008.12.009

VE rose from the empirical observation that many patients perceived these feelings in the months preceding a cardiac event [4]. Subsequently conducted studies showed that VE predicted first-time myocardial infarction [5] and recurrent cardiac events after coronary angioplasty [6]. In terms of depression, the relationship with incident CAD was shown to lie along a continuum with, as yet, minimal symptoms of depression increasing CAD risk [3,7]. The psychophysiological mechanisms in the relationship between VE, depression, and CAD are not fully understood [1,2,4] but could partially relate to a hypercoagulable state that is characterized by activated coagulation and/or impaired fibrinolysis [8]. For instance, clotting factor VII,

94

R. von Känel et al. / Journal of Psychosomatic Research 67 (2009) 93–101

fibrinogen, and the main fibrinolysis inhibitor plasminogen activator inhibitor-1 were all elevated in VE [9–11] and depression [12–14]. Chronic low-grade hypercoagulability promotes atherosclerosis and its thrombotic complications [15,16]. Acute psychosocial stress physiologically enhances coagulation activity [8] which, if exaggerated, might increase the risk of atherothrombotic events in susceptible individuals, such as in those with negative affect, including depression [17,18]. As opposed to individual clotting and fibrinolysis factors, the hypercoagulability marker fibrin D-dimer indicates overall activation of coagulation and fibrinolysis. D-dimer is generated when the fibrinolytic enzyme plasmin dissolves fibrin—the end product of the coagulation cascade and main component of a blood clot [19]. A meta-analysis established the role of D-dimer as a predictor of CAD in healthy individuals and in patients with pre-existing vascular disease [20]. D-dimer showed reliable increases to acute psychosocial stress across different studies and populations [21–24]. We hypothesized an association of VE and depressive symptoms with changes in D-dimer in response to and during recovery from acute psychosocial stress in German teachers who are a population with compromised physical and mental health [25]. In a secondary analysis, we covaried for sex, age, and reactivity of stress hormones and blood pressure (BP) because these variables might affect the hypothesized relationships. Specifically, resting D-dimer is higher in women than in men [26] and it increases with age [24]. Norepinephrine (NE) and BP reactivity directly correlated with immediate increase in D-dimer and changes in other procoagulant molecules to acute psychosocial stress [23,27]. Depressive symptoms were associated with enhanced NE and cardiovascular stress reactivity [28]. Vital exhaustion was associated with dampened cortisol stress reactivity [29,30] which, in turn, might elicit enhanced coagulation reactivity [31]. Finally, a mediational analysis was performed to explore whether depression would mediate the effect of exhaustion on D-dimer change or vice versa given that the two constructs share conceptual similarity [4].

Materials and methods Study design and participants The present study is part of the Trier Teacher Stress Study investigating mental and physical health in school teachers from Germany and Luxembourg as previously described [32,33]. Specific exclusion criteria for enrollment were diabetes, a positive history of cancer, artery disease, heart failure, medication with corticosteroids, psychotropics, antidepressants, and pregnancy. We also excluded participants with any psychiatric disorder, including clinical forms of depressive disorders, and those who had previously undergone treatment for any depressive disorder. For the present study, 42 employed teachers consented to undergo

twice an acute psychosocial stressor (cf. below). On the first test day, sex and age were recorded and all subjects completed psychological questionnaires (cf. below). We excluded two subjects who were smokers and one woman with oral contraceptive use, as these conditions may affect D-dimer levels [34,35]. We further excluded one subject with incomplete D-dimer measures, leaving a final sample of 38 teachers for the analysis. The ethic committees of the State Medical Association of Rheinland Pfalz and the University of Trier approved the study protocol. All participants provided written informed consent and were paid €70 on completion of the stress protocol. Assessment of psychological measures Vital exhaustion We measured VE using a German version of the nine-item short form [11] of the original 21-item Maastricht VE Questionnaire [36]. Scores of the short version correlate well (r=.94) with those from the original 21-item questionnaire [37]. Items ask about undue fatigue, disturbed sleep, general malaise, irritability, loss of mental and physical energy, and feelings of demoralization. Typical items are “Do you often feel tired?,” “Do you feel weak all over?,” and “Do you sometimes feel that your body is like a battery that is losing its power?” [37]. Possible answers to each item are ‘‘no’’ (score=0), “uncertain” (score=1), and ‘‘yes’’ (score=2), giving rise to a total VE score between 0 and 18. Scores from 0 to 3 indicate “no VE,” scores from 4 to 10 are equivalent to “mild to moderate VE,” scores from 11 to 14 reflect “substantial VE,” and scores N14 indicate “severe VE”. Depression We applied the validated German version [38] of the depression subscale of the original Hospital Anxiety and Depression Scale (HADS-D) designed to rate symptom levels of depression in non-psychiatric populations [39]. Each of its seven items is rated on a four-point Likert scale (0=not at all, 3=mostly) giving rise to a total depressive symptom score between 0 and 21. Total scores are clinically interpreted with the following cut-off points: 0–7: no depression; 8–10: mild depression; 11–14: moderate depression; 15–21: severe depression. The HADS-D has excellent psychometric properties showing a test–retest reliability of 0.92 in healthy respondents [40]. Cronbach's alpha for the English and German versions of the HADS-D range between 0.81 and 0.90 [41]. Stress experiment Subjects were assessed between 3 and 4 p.m. on two separate days with a maximum interval of four days (i.e., no weekend in between assessments) and instructed to refrain from exercise, a heavy lunch and alcoholic drinks on the test day. Premenopausal women were tested during the luteal phase of the menstrual cycle to avoid confounding of the

R. von Känel et al. / Journal of Psychosomatic Research 67 (2009) 93–101

procoagulant stress response by female sex hormones [42]. A venous catheter was placed to the dominant forearm. After a 50-min rest and information on the stress protocol, subjects underwent the Trier Social Stress Test (TSST). The TSST consists of a 3-min preparation phase, a 5-min job interview, and a 5-min mental arithmetic task all in front of an audience [43,44] and elicits firm increases in D-dimer, NE, cortisol, and blood pressure (BP) [23,31]. Sampling of blood, saliva, and blood pressure The different times for the assessment of D-dimer, stress hormones, and BP were selected to comply with protocols of our former stress studies, in which the respective time points accurately tracked changes over time in these measures [23,24,31,45]. Blood for D-dimer measures was collected immediately before stress, immediately after cessation of stress, and 20, 45, and 90 min after stress cessation. Blood for NE measures was collected immediately before stress, immediately after cessation of stress, and 10 and 20 min after stress cessation. For the present analysis, saliva for free cortisol measures was obtained immediately before stress, and 10, 20, and 30 min after cessation of stress. Systolic and diastolic BP was measured immediately before stress, immediately after stress cessation, and 10 min after cessation of stress using an automated device (Omron M5-I, Mannheim, Germany). The highest value of NE, cortisol, and mean arterial pressure (MAP) measured at any time point post-stress was defined as the peak level of each subject. Stress reactivity of NE, cortisol, and MAP was calculated by subtracting pre-stress levels from peak levels. D-dimer, NE, cortisol, and BP measures were all collected on Days 1 and 2. For the statistical analysis, we calculated and used the average of Days 1 and 2 of resting and reactivity levels in order to enhance the robustness of individual differences in the biological stress response [46]. Biochemical analysis For D-dimer measurement, 3.0 ml of venous blood was added to 0.3 ml of citrate (0.106 Mol/L) monovettes (Sarstedt, Nümbrecht, Germany). Samples were immediately centrifuged at 4°C for 15 min at 2000×g. Within 1 h, plasma aliquots were transferred to the core laboratory (Synlab, Trier, Germany) and processed immediately. D-dimer was determined by means of an ELISA kit (VIDAS D-dimer Exclusion TM, bioMérieux Marcy-l'Etoile, France). Intraand interassay coefficients of variation (CV) were b6.3%. Saliva was collected in 2 ml reaction tubes (Sarstedt, Nümbrecht, Germany) and stored at −20°C until analysis. Salivary free cortisol was measured by an in-house timeresolved fluorescence immunoassay (DELFIA, intra- and interassay CV b11.6%). For NE measurements, venous blood was sampled in EDTA monovettes and centrifuged for 15 min at 2000×g at 4°C. Plasma samples were immediately stored at −80°C and analyzed by the core lab using high-

95

performance liquid chromatography (intra- and interassay CV b10%). Statistical analysis Data were analyzed using SPSS 15.0 for Windows (SPSS, Chicago, IL, USA) with level of significance at P≤.05 (twotailed). All data showed a normal distribution as verified by the Kolmogorov–Smirnov test, except D-dimer measures, which were logarithmically transformed before analysis. For clarity, D-dimer values are presented in original units. Independent t tests and chi-square tests were used to estimate group differences in continuous and categorical variables, respectively. Pearson correlation analysis was applied to estimate the bivariate correlation coefficient between two variables controlling for covariates. For illustrative purposes, we performed a median split on VE and depression scores (“high” vs. “low” scores) according to which we presented subject characteristics and stress-induced D-dimer changes. As a means of data reduction, we calculated mean arterial pressure (MAP) [2/3(diastolic BP)+1/3(systolic BP)] to be used in statistical analysis. We applied repeated-measures analysis with covariates sex and age to investigate whether the stress-induced change in D-dimer levels across the five time points (i.e., “time effect”) was continuously associated with continuous measures of psychological scores (i.e., “time-by-psychological score interaction”). In a next step, we additionally forced reactivity measures of stress hormones and MAP into the model. The Huynh–Feldt correction for the degrees of freedom was applied whenever the sphericity assumption was violated. Effect sizes are expressed as partial eta squared (ηp2 ) interpreted as the percentage of variance in the dependent variable uniquely attributable to the independent variable. ηp2 Values of .01, .06, and .14, respectively, are considered small, medium, and large effects [47]. A significant association of the depression and VE score with stress-induced changes in D-dimer levels over time can be considered a significant omnibus test. We thus explored post hoc which items of the VE and HADS-D scales individually contributed to this association without making adjustment for multiple testing to generate hypotheses for future studies. The Sobel test was used to test for a mediating effect [48], one in which a variable is assumed to explain how or why an effect occurs [49]. Specifically, we tested whether VE mediates the effect of depression, or vice versa, on stressinduced D-dimer changes.

Results Subject characteristics Table 1 presents the characteristics of the 17 male and 21 female teachers studied and their VE and depression category based on a median split of respective scores. Vital exhaustion

96

R. von Känel et al. / Journal of Psychosomatic Research 67 (2009) 93–101

Table 1 Characteristics of subjects by vital exhaustion and depression categorization

Women/men [%] Age (years) VE score Depression score SBP (mm Hg) DBP (mm Hg) NE (pg/ml) Cortisol (nmol/l) D-dimer (ng/ml)

All subjects (n=38)

High vital exhaustion (n=19)

Low vital exhaustion (n=19)

P

High depression (n=18)

Low depression (n=20)

P

55/45 50±8 7.2±5.1 4.1±3.5 124±19 81±10 474±140 3.4±1.5 391±248

53/47 52±7 11.5±2.8 5.4±3.1 127±19 82±9 458±120 3.0±1.3 430±257

58/42 48±10 2.9±2.5 2.9±3.6 121±19 80±11 490±160 3.7±1.6 351±239

.744 .124 b.001 .029 .306 .550 .493 .147 .140

44/56 50±9 9.5±4.7 7.0±3.0 127±16 83±9 494±101 3.4±2.5 350±114

65/35 51±9 5.1±4.6 1.6±1.2 121±21 79±10 455±169 3.3±1.5 427±327

.203 .696 .006 b.001 .306 .200 .389 .900 .788

Values are given as mean±SD and in original units. Measures of systolic blood pressure, diastolic blood pressure, norepinephrine, cortisol, and D-dimer are the average of pre-stress values of the two test days. SBP, systolic blood pressure; DBP, diastolic blood pressure.

and depression correlated significantly in all subjects (r=.47, P=.003). Based on clinical cutoffs, 9 subjects had no VE, 18 had mild to moderate VE, 8 had substantial VE, and 3 had severe VE. Five subjects had clinically relevant levels of depressed mood (score N8). Pre-stress D-dimer levels (i.e., average of the two test days) did not significantly correlate with any variable shown in Table 1. Neuroendocrine and hemodynamic reactivity The increase from pre-stress to peak levels was significant (P values b.001) for cortisol (3.4±1.5 vs. 6.1±3.6 nmol/l), NE (474±140 vs. 737±214 pg/ml), and MAP (95±12 vs. 106±13 mmHg). Controlling for sex and age, depression correlated with cortisol reactivity (r=−.37, P=.026), but not with reactivity of NE (r=−.19, P=.26) and of MAP (r=−.16, P=.37). Vital exhaustion showed no significant correlation with reactivity of cortisol (r=−.14, P=.43), NE (r=−.22, P=.20), and MAP (r=−.18, P=.32), taking into account sex and age. Vital exhaustion, depression, and stress-induced D-dimer change There was an interaction between time and VE (F4,136=2.97, P=.022; η2=.080) and between time and

depression (F4,136=3.38, P=.011; η2=.090) for the stressinduced change in D-dimer levels over time controlling for sex (P values ≥.28) and age (P values ≥.24). Table 2 shows that elevated levels of VE and depression were significantly correlated with less D-dimer increase between pre-stress and immediately post-stress on the one hand and with greater D-dimer increase between 20 and 45 min poststress on the other. Fig. 1 illustrates that, based on a median split on psychological scores, subjects with low VE and with low depression, respectively, had a greater D-dimer increase from pre-stress to post-stress than subjects scoring high on VE and depression ([1] in Fig. 1). Whereas subjects with high VE and high depression, respectively, experienced D-dimer increase between 20 and 45 min post-stress, those who scored low on VE and depression, respectively, showed a decrease in D-dimer during this time interval ([2] in Fig. 1). Role of neuroendocrine and hemodynamic reactivity We computed a series of analyses to investigate whether the association between D-dimer change across time points and VE and depression was accounted for by stress hormones and MAP reactivity. The subsequent analyses all controlled for sex and age.

Table 2 Partial correlations between psychological scores and D-dimer changes Time intervals for D-dimer measurements Scores of scales and items

Immediately post-stress minus pre-stress

20 min post-stress minus immediately post-stress

45 min post-stress minus 20 min post-stress

90 min post-stress minus 45 min post-stress

Vital exhaustion scale Often feel tired Depression scale Future enjoyment Leisure time enjoyment

−.46 ⁎⁎ −.42 ⁎ −.51 ⁎⁎ −.51 ⁎⁎ −.34 ⁎

−.02 −.05 .12 .22 −.03

.34 ⁎ .17 .41 ⁎ .47 ⁎⁎ .40 ⁎

.06 .20 −.13 −.30 −.04

The columns show the partial correlations coefficients (r) adjusted for sex and age. ⁎ Pb.05. ⁎⁎ Pb.01.

R. von Känel et al. / Journal of Psychosomatic Research 67 (2009) 93–101

97

Cortisol reactivity The interaction between time and VE (F4,132=2.39, Pb.06; η 2 =.067) and between time and depression (F4,132=2.19, Pb.08; η2=.062) became marginally significant when controlling for cortisol reactivity. Cortisol reactivity did not interact with VE (P=.75) and depression (P=.79) in determining D-dimer change over time. Norepinephrine reactivity Controlling for NE reactivity maintained the significance of the relationships between stress-induced changes in D-dimer over time and VE (F4,132=2.64, P=.037; η2=.074) and depression (F4,132=3.43, P=.011; η2=.094). Norepinephrine reactivity did not interact with VE (P=.13) and depression (P=.46) in determining D-dimer change over time. Blood pressure reactivity Controlling for MAP reactivity maintained the significance of the relationships between stress-induced changes in D-dimer over time and VE (F4,124=3.06, P=.019; η2=.090) and depression (F4,124=3.85, P=.006; η2=.011). Reactivity of MAP did not interact with VE (P=.89) and depression (P=.96) in determining D-dimer change over time. Individual item analysis We performed an individual item analysis to possibly identify the specific components of the VE and depression constructs contributing to the stress-induced change in D-dimer levels over time with covariates sex and age. A significant time-by-item interaction was found for “Do you often feel tired?” (F4,136=2.44, P=.050; η2=.067) from the VE scale, and for “I look forward with enjoyment to things” (F4,136=4.36, P=.002; η2=.114) and “I can enjoy a good book or TV program” (F4,136=2.54, P=.043; η2=.069) from the HADS-D scale. Table 2 shows that elevated levels of “feeling tired”, “not looking forward with enjoyment to things”, and “not enjoying leisure time activity” were all significantly associated with less increase in D-dimer from immediately pre-stress to immediately post-stress. Elevated levels of “not looking forward with enjoyment to things” and of “not enjoying leisure time activity” were also significantly associated with a greater D-dimer increase between 20 and 45 min post-stress. Combined effect of vital exhaustion and depression on stress-induced D-dimer change Fig. 1. Association between vital exhaustion and depression with D-dimer change over time due to acute psychosocial stress. Panels A and B illustrate the stress-induced D-dimer change between immediately pre-stress (pre-str), immediately post-stress (post-str), 20, 45, and 90 min post-stress for subjects with high vital exhaustion (score 8-16; n=19) vs. those with low vital exhaustion (score 0-7; n=19) and for subjects with high levels of depression (score 4-15; n=18) vs. those with low levels of depression (score 0-3; n=20) as per a median split of respective scores. Adjustment was made for sex and age. D-dimer values are given as means±S.E.M. and in original units.

Because VE and depression scores were significantly associated with each other and with D-dimer change over time, we explored to what extent their procoagulant effect would be independent of each other or, alternatively, relate to each other. Vital exhaustion (P=.32) and depression (P=.19) were not associated with stress-induced D-dimer change over time when entering the model simultaneously

98

R. von Känel et al. / Journal of Psychosomatic Research 67 (2009) 93–101

suggesting their separate effects partialled out each other. Also, VE and depression scores did not interact in determining D-dimer change across time points (P=.73), suggesting that VE and depression did not moderate each other's effects. Vital exhaustion was not revealed to be a mediator of the association between depression and immediate D-dimer increase (P=.19) or between depression and D-dimer change between 20 and 45 min post-stress (P=.38). In turn, depression was not revealed to be a mediator of the association between VE and immediate D-dimer increase (P=.10) or between VE and change in D-dimer between 20 and 45 min post-stress (P=.13).

Discussion We found that a state of vital exhaustion and depressive symptoms were individually associated with stress-induced changes over time in the coagulation activation marker fibrin D-dimer in healthy, nonsmoking school teachers independent of sex and age. The two psychological constructs showed similar associations with stress-induced D-dimer change. Specifically, higher levels of VE and of depressed mood were both associated with reduced D-dimer increase between pre-stress and immediately post-stress and with higher D-dimer during 20 and 45 min into recovery from stress. This observation concurs with the concept of a perturbed biological stress response as characterized by a dampened immediate reaction of the coagulation system to stress on the one hand and a delayed recovery of stressinduced coagulation changes to pre-stress levels of activity on the other [50]. Normal adaptive physiology posits a brisk increase in coagulation activity to acute stress in order to protect the organism from bleeding in fight-flight situations [17]. This physiological hypercoagulability typically restores within the first 20 min after stress [31]. A procoagulant milieu promotes coronary thrombus formation in the hours after coronary plaque disruption [16]. Therefore, stressinduced hypercoagulability, which in relation to increased VE and depression carries over to a critically long period after stress cessation, might enhance atherothrombotic risk [51]. Notably, such reasoning seems of clinical significance in that effects were medium to large with VE and depression accounting for 8% and 9%, respectively, of the variance in D-dimer change over time. Taken together, the observed associations support the notion that acute psychosocial stress and negative affect could interact in increasing the risk of CAD and its thrombotic complications by kindling a hypercoagulable state [1–3,18]. It should be mentioned that D-dimer formation is a complex process that involves several coagulation and fibrinolysis steps. The molar concentration of D-dimer is regulated by plasmin generation, inhibition of plasmin by antiplasmin, plasmin degradation of fibrin, and hepatic clearance of D-dimer [52], all of which may vary in their

kinetics depending upon study protocols [23]. In a pig model of venous thrombosis, an exogenous bolus of fibrinolysis-stimulating tissue plasminogen activator resulted in a D-dimer peak response 30 min later, whereby D-dimer had not returned to resting levels 2 1/2 h after bolus injection [53]. This could partially be explained by the observation that D-dimer is cleared from the circulation with an average half-life of 7 h [52]. Therefore, what we interpret as delayed recovery of D-dimer from stress in relation to both VE and depression could also reflect a delayed increase in D-dimer towards its peak response. However, we could have only resolved this if we had extended sampling of D-dimer beyond 90 min post-stress. In order to generate hypotheses for future studies, we performed a post hoc analysis using individual items related to the VE and depression constructs. Tiredness on the one hand and depressive symptoms related to anhedonia on the other contributed most to the association of VE and depression with D-dimer change over time. As seen for the total symptom scores of the VE and HADS-D scales, elevated levels of “often feel tired”, “not looking forward with enjoyment to things”, and “not enjoying reading a book or watching TV” were associated with blunted increase in D-dimer levels from immediately pre-stress to immediately post-stress. Also, elevated levels of tiredness and anhedonia were both positively associated with D-dimer change between 20 and 45 min post-stress implying compromised recovery of D-dimer during this time interval. Tiredness and related fatigability are viewed as somatic symptoms of depression, whereas anhedonia reflects depressed cognition [54]. Recent research suggests that somatic and cognitive symptoms of depression might affect the course of CAD differently, with somatic symptoms perhaps exerting greater impact [55]. Our study did not distinguish different effects of somatic as opposed to cognitive symptoms on stress hypercoagulability. Future research may want to apply validated psychometric instruments that include somatic and cognitive dimensions of depression to sizeable samples allowing for factor analytic approaches. We found that the individual associations of VE and depression with stress-induced D-dimer changes were not independent of each other. Specifically, individual effects became nonsignificant when VE and depression scores entered the equation together. There is a debate about the extent to which VE and depression reflect distinct constructs [56]. In a working population from Germany similar in socioeconomic status and cultural background to the teachers investigated here, we previously performed a factor analysis. We found that while the short form of the VE questionnaire and the HADS-D subscale showed some conceptual overlap, the scales constituted distinct psychological concepts [57]. Specifically, it has been proposed that feelings of guilt, hopelessness, and worthlessness are rather unique to the conceptualization of depression, whereas lack of energy, fatigue, and absence of sadness and of cognitive distortions are characteristic for VE [4,56,58]. In terms of effects on

R. von Känel et al. / Journal of Psychosomatic Research 67 (2009) 93–101

acute stress hypercoagulability, our data suggest that the degree of conceptual overlap between VE and depression was substantial. This might also explain why the two concepts were not in a statistical sense moderator and mediator of each other's effect on stress-induced D-dimer change across all time points. We found little evidence for the assumption that stress hormone and cardiovascular reactivity affected the association between VE and depression on the one hand and stressinduced D-dimer change on the other. Covarying for cortisol reactivity, but not for NE reactivity and MAP reactivity, rendered these relationships marginally significant. However, effect sizes of VE and depression were kept at moderate levels and a decrease in power due to a reduction in degrees of freedom, needs to be considered when interpreting this finding. Our findings are further to be interpreted within the limitations of the study design. We recruited middle-aged, apparently healthy teachers at high functional level and, on average, with low levels of depressive mood. As a consequence of the latter, the range of scores for the HADS-D scale was likely truncated, potentially leading to lower correlations between depression and the other variables. Therefore, our data cannot be generalized to elderly individuals and patients with cardiovascular disease and clinical depression which are populations running a comparably higher risk of atherothrombotic events. To avoid overfitting of statistical models, the relatively small sample size did not allow us to statistically control for covariates above and beyond sex and age. Particularly, we are unable to exclude the possibility that life style and metabolic factors might account for some of the observed relationships. For instance, dyslipidemia and elevated body mass index previously correlated with increased D-dimer levels [59,60]. Also, while reactivity measures did not substantially affect relationships individually, when aggregated and applied to a larger sample, they might account for some of the effects observed. We did not consider measures of positive affect to demonstrate discriminative validity of VE and depression [61] and also not measures of negative affect above and beyond VE and depression. Particularly, hostility and anxiety were associated with the procoagulant stress response in previous studies [62,63]. Therefore, it is possible that VE and depression are not unique predictors of D-dimer response in our study participants and do also not have discriminative validity in this regard. Taken together, our findings suggest a blunted immediate D-dimer stress response and delayed recovery of D-dimer levels from stress with elevated levels of VE and depressive symptoms, even at a subthreshold level of clinical depression. Tiredness and anhedonia were important in this relationship. The prolonged hypercoagulability after stress cessation might particularly contribute to the atherothrombotic risk previously observed in individuals with feelings of vital exhaustion and depression.

99

Acknowledgments This work was supported by Emmy Noether research grant KU 1401/4-1, KU 1401/4-2, and KU 1401/4-3 of the German Research Foundation (DFG) awarded to B.M.K. B.M.K. and S.B. are members of the International Research Training Group IRTG funded by the DFG (GRH 1389/1). References [1] Rozanski A, Blumenthal JA, Kaplan J. Impact of psychological factors on the pathogenesis of cardiovascular disease and implications for therapy. Circulation 1999;99:2192–217. [2] Melamed S, Shirom A, Toker S, Berliner S, Shapira I. Burnout and risk of cardiovascular disease: evidence, possible causal paths, and promising research directions. Psychol Bull 2006;132:327–53. [3] Rugulies R. Depression as a predictor for coronary heart disease. A review and meta- analysis. Am J Prev Med 2002;23:51–61. [4] Appels A. Exhaustion and coronary heart disease: the history of a scientific quest. Patient Educ Couns 2004;55:223–9. [5] Appels A, Mulder P. Excess fatigue as a precursor of myocardial infarction. Eur Heart J 1988;9:758–64. [6] Kop WJ, Appels AP, Mendes de Leon CF, de Swart HB, Bär FW. Vital exhaustion predicts new cardiac events after successful coronary angioplasty. Psychosom Med 1994;56:281–7. [7] Wulsin LR, Evans JC, Vasan RS, Murabito JM, Kelly-Hayes M, Benjamin EJ. Depressive symptoms, coronary heart disease, and overall mortality in the Framingham Heart Study. Psychosom Med 2005;67:697–702. [8] von Känel R, Mills PJ, Fainman C, Dimsdale JE. Effects of psychological stress and psychiatric disorders on blood coagulation and fibrinolysis: a biobehavioral pathway to coronary artery disease? Psychosom Med 2001;63:531–44. [9] von Känel R, Mausbach BT, Kudielka BM, Orth-Gomér K. Relation of morning serum cortisol to prothrombotic activity in women with stable coronary artery disease. J Thromb Thrombolysis 2008;25: 165–72. [10] Kudielka BM, Bellingrath S, von Känel R. Circulating fibrinogen but not D-dimer level is associated with vital exhaustion in school teachers. Stress 2008;11:250–8. [11] von Känel R, Frey K, Fischer J. Independent relation of vital exhaustion and inflammation to fibrinolysis in apparently healthy subjects. Scand Cardiovasc J 2004;38:28–32. [12] Doulalas AD, Rallidis LS, Gialernios T, Moschonas DN, Kougioulis MN, Rizos I, Tselegaridis TS, Kremastinos DT. Association of depressive symptoms with coagulation factors in young healthy individuals. Atherosclerosis 2006;186:121–5. [13] Kop WJ, Gottdiener JS, Tangen CM, Fried LP, McBurnie MA, Walston J, Newman A, Hirsch C, Tracy RP. Inflammation and coagulation factors in persons N65 years of age with symptoms of depression but without evidence of myocardial ischemia. Am J Cardiol 2002;89:419–24. [14] Lahlou-Laforet K, Alhenc-Gelas M, Pornin M, Bydlowski S, Seigneur E, Benetos A, Kierzin JM, Scarabin PY, Ducimetiere P, Aiach M, Guize L, Consoli SM. Relation of depressive mood to plasminogen activator inhibitor, tissue plasminogen activator, and fibrinogen levels in patients with versus without coronary heart disease. Am J Cardiol 2006;97:1287–91. [15] Falk E, Fernández-Ortiz A. Role of thrombosis in atherosclerosis and its complications. Am J Cardiol 1995;75:3B–11B. [16] Libby P. The molecular mechanisms of the thrombotic complications of atherosclerosis. J Intern Med 2008;263:517–27. [17] von Känel R. Hemostasis and Stress. In: Fink G, editor. Encyclopedia of stress. 2 ed. Vol 2. Oxford: Elsevier Academic Press, 2007. pp. 300–5.

100

R. von Känel et al. / Journal of Psychosomatic Research 67 (2009) 93–101

[18] Bhattacharyya MR, Steptoe A. Emotional triggers of acute coronary syndromes: strength of evidence, biological processes, and clinical implications. Prog Cardiovasc Dis 2007;49:353–65. [19] Lip GY, Lowe GD. Fibrin D-dimer: a useful marker of thrombogenesis? Clin Sci 1995;89:205–14. [20] Danesh J, Whincup P, Walker M, Lennon L, Thomson A, Appelby P, Rumley A, Lowe GD. Fibrin D-dimer and coronary heart disease: prospective study and meta-analysis. Circulation 2001;103:2323–7. [21] von Känel R, Mills PJ, Ziegler MG, Dimsdale JE. Effect of beta2adrenergic receptor functioning and increased norepinephrine on the hypercoagulable state with mental stress. Am Heart J 2002;144:68–72. [22] von Känel R, Dimsdale JE, Patterson TL, Grant I. Acute procoagulant stress response as a dynamic measure of allostatic load in Alzheimer caregivers. Ann Behav Med 2003;26:42–8. [23] Wirtz PH, Ehlert U, Emini L, Rüdisüli K, Groessbauer S, Gaab J, Elsenbruch S, von Känel R. Anticipatory cognitive stress appraisal and the acute procoagulant stress response in men. Psychosom Med 2006; 68:851–8. [24] Wirtz PH, Redwine LS, Baertschi C, Spillmann M, Ehlert U, von Känel R. Coagulation activity before and after acute psychosocial stress increases with age. Psychosom Med 2008;70:476–81. [25] Weber A, Weltle D, Lederer P. Ill health and early retirement among school principals in Bavaria. Int Arch Occup Environ Health 2005;78: 325–31. [26] Rudnicka AR, Rumley A, Lowe GD, Strachan DP. 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 2007;115: 996–1003. [27] Steptoe A, Kunz-Ebrecht S, Rumley A, Lowe GD. Prolonged elevations in haemostatic and rheological responses following psychological stress in low socioeconomic status men and women. Thromb Haemost 2003;89:83–90. [28] Light KC, Kothandapani RV, Allen MT. Enhanced cardiovascular and catecholamine responses in women with depressive symptoms. Int J Psychophysiol 1998;28:157–66. [29] Kristenson M, Orth-Gomér K, Kucinskienë Z, Bergdahl B, Calkauskas H, Balinkyniene I, Olsson AG. Attenuated cortisol response to a standardized stress test in Lithuanian versus Swedish men: the LiVicordia study. Int J Behav Med 1998;5:17–30. [30] Nicolson NA, van Diest R. Salivary cortisol patterns in vital exhaustion. J Psychosom Res 2000;49:335–42. [31] Wirtz PH, Ehlert U, Emini L, Rüdisüli K, Groessbauer S, Mausbach BT, von Känel R. The role of stress hormones in the relationship between resting blood pressure and coagulation activity. J Hypertens 2006;24:2409–16. [32] Bellingrath S, Weigl T, Kudielka BM. Cortisol dysregulation in school teachers in relation to burnout, vital exhaustion, and effort-rewardimbalance. Biol Psychol 2008;78:104–13. [33] von Känel R, Bellingrath S, Kudielka BM. Association between burnout and circulating levels of pro- and anti-inflammatory cytokines in school teachers. J Psychosom Res 2008;65:51–9. [34] Yanbaeva DG, Dentener MA, Creutzberg EC, Wesseling G, Wouters EF. Systemic effects of smoking. Chest 2007;131:1557–66. [35] Archer DF, Mammen EF, Grubb GS. The effects of a low-dose monophasic preparation of levonorgestrel and ethinyl estradiol on coagulation and other hemostatic factors. Am J Obstet Gynecol 1999; 181:63–6. [36] Appels A, Höppener P, Mulder P. A questionnaire to assess premonitory symptoms of myocardial infarction. Int J Cardiol 1987;17:15–24. [37] Kopp MS, Falger PR, Appels A, Szedmak S. Depressive symptomatology and vital exhaustion are differentially related to behavioral risk factors for coronary artery disease. Psychosom Med 1998;60:752–8. [38] Herrmann C, Buss U, Snaith RP. HADS-D Hospital Anxiety and Depression Scale Deutsche Version. Ein Fragebogen zur Erfassung von Angst und Depressivität in der somatischen Medizin. Bern: Verlag Hans Huber, 1995.

[39] Zigmond AS, Snaith RP. The Hospital Anxiety and Depression Scale. Acta Psychiatr Scand 1983;67:361–70. [40] Snaith RP, Zigmond AS. The Hospital and Depression Scale Manual. Windsor, Berkshire, England: Nfer-Nelson, 1994. [41] Herrmann C. International experiences with the Hospital Anxiety and Depression Scale: a review of validation data and clinical results. J Psychosom Res 1997;42:17–41. [42] Jern C, Manhem K, Eriksson E, Tengborn L, Risberg B, Jern S. Hemostatic responses to mental stress during the menstrual cycle. Thromb Haemost 1991;66:614–8. [43] Kirschbaum C, Pirke KM, Hellhammer DH. The ‘Trier Social Stress Test’—a tool for investigating psychobiological stress responses in a laboratory setting. Neuropsychobiology 1993;28:76–81. [44] Kudielka BM, Hellhammer DH, Kirschbaum C. Ten years of research with the Trier Social Stress Test (TSST)—revisited. In: Harmon-Jones E, Winkielman P, editors. Social Neuroscience: integrating biological and psychological explanations of social behavior. New York: Guilford Press, 2007. pp. 56–83. chapter 4. [45] von Känel R, Preckel D, Zgraggen L, Mischler K, Kudielka BM, Haeberli A, Fischer JE. The effect of natural habituation on coagulation responses to acute mental stress and recovery in men. Thromb Haemost 2004;92:1327–35. [46] Kamarck TW, Debski TT, Manuck SB. Enhancing the laboratory-tolife generalizability of cardiovascular reactivity using multiple occasions of measurement. Psychophysiology 2000;37:533–42. [47] Cohen J. Statistical power analysis for the behavioral sciences. New York: Lawrence Erlbaum Associates, 1997. [48] Sobel ME. Asymptotic intervals for indirect effects in structural equations models. In: Leinhart S, editor. Sociological methodology. San Francisco: Jossey-Bass, 1982. pp. 290–312. [49] Kraemer HC, Stice E, Kazdin A, Offord D, Kupfer D. How do risk factors work together? Mediators, moderators, and independent, overlapping, and proxy risk factors. Am J Psychiatry 2001;158: 848–56. [50] McEwen BS. Protective and damaging effects of stress mediators. N Engl J Med 1998;338:171–9. [51] Wirtz PH, Ehlert U, Emini L, Rüdisüli K, Groessbauer S, von Känel R. Procoagulant stress reactivity and recovery in apparently healthy men with systolic and diastolic hypertension. J Psychosom Res 2007;63: 51–8. [52] Chandler WL, Velan T. Plasmin generation and D-dimer formation during cardiopulmonary bypass. Blood Coagul Fibrinolysis 2004;15: 583–91. [53] Robinson VJ, Pineda GE, Salah AK, Pipkin WL, Corley JH, Jonah MH, Gossage JR. Latex D-dimer signal in in situ femoral vein thrombus in swine and effect of minidose exogenous tissue plasminogen activator bolus. Chest 2005;127:622–9. [54] Beck AT, Steer RA. Manual for the revised Beck Depression Inventory. San Antonio, Texas: Psychological Corporation, 1987. [55] De Jonge P, Ormel J, van den Brink RH, van Melle JP, Spijkerman TA, Kuijper A, van Veldhuisen DJ, van den Berg MP, Honig A, Crijns HJ, Schene AH. Symptom dimensions of depression following myocardial infarction and their relationship with somatic health status and cardiovascular prognosis. Am J Psychiatry 2006; 163:138–44. [56] van Diest R, Appels A. Vital exhaustion and depression: a conceptual study. J Psychosom Res 1991;35:535–44. [57] Kudielka BM, von Känel R, Gander ML, Fischer JE. The interrelationship of psychosocial risk factors for coronary artery disease in a working population: do we measure distinct or overlapping psychological concepts? Behav Med 2004;30:35–43. [58] Kop WJ, Hamulyák K, Pernot C, Appels A. Relationship of blood coagulation and fibrinolysis to vital exhaustion. Psychosom Med 1998; 60:352–8. [59] Hughes R, Thomson K, Hopkins R, Weatherall M, Wiltshire C, Wilsher M, Beasley R. Determinants of plasma D-dimer levels in a traveling population. J Thromb Haemost 2005;3:2445–8.

R. von Känel et al. / Journal of Psychosomatic Research 67 (2009) 93–101 [60] Gallistl S, Sudi KM, Borkenstein M, Weinhandl G, Zotter H, Muntean W. Correlation between cholesterol, soluble P-selectin, and D-dimer in obese children and adolescents. Blood Coagul Fibrinolysis 2000;11: 755–60. [61] Chida Y, Steptoe A. Positive psychological well-being and mortality: a quantitative review of prospective observational studies. Psychosom Med 2008;70:741–56.

101

[62] Markovitz JH, Matthews KA, Kiss J, Smitherman TC. Effects of hostility on platelet reactivity to psychological stress in coronary heart disease patients and in healthy controls. Psychosom Med 1996;58:143–9. [63] von Känel R, Dimsdale JE, Adler KA, Patterson TL, Mills PJ, Grant I. Effects of depressive symptoms and anxiety on hemostatic responses to acute mental stress and recovery in the elderly. Psychiatry Res 2004; 126:253–64.