Depressive symptoms and attenuated physiological reactivity to laboratory stressors

Biological Psychology 87 (2011) 430–438

Contents lists available at ScienceDirect

Biological Psychology journal homepage: www.elsevier.com/locate/biopsycho

Depressive symptoms and attenuated physiological reactivity to laboratory stressors Andreas Schwerdtfeger ∗ , Ann-Kathrin Rosenkaimer Johannes Gutenberg-University Mainz, Germany

a r t i c l e

i n f o

Article history: Received 1 June 2010 Accepted 30 May 2011 Available online 14 June 2011 Keywords: Cardiovascular reactivity Depressive symptoms Electrodermal reactivity Motivational deficit Self-relevant stressors

a b s t r a c t There is evidence that depressive symptoms are associated with attenuated physiological reactivity to active stressors. However, it is not known whether blunted reactivity in depressed individuals is stressor-specific. We examined cardiovascular and electrodermal reactivity in non-clinical participants with varying levels of depressive symptoms to different active and passive stressors. Depressive symptoms were inversely related to both blood pressure and skin conductance reactivity during a public speaking task and the viewing of the speech video. However, no effects were found during a cold pressor task. Together these findings suggest that depressive symptoms are related to attenuated sympathetic nervous system reactivity in response to self-relevant stressors. © 2011 Elsevier B.V. All rights reserved.

1. Introduction Depression has been discussed to constitute a robust psychosocial risk factor for cardiovascular diseases (CVD; Barth et al., 2004; Rugulies, 2002; Wulsin and Singal, 2003). In particular, individuals with depressive symptoms are at higher risk for developing coronary artery disease, myocardial infarction, and complications following heart surgery. It has been suggested that the relationship between depression and CVD is mediated by behavioral factors on the one hand (e.g., substance abuse, diminished physical activity), and dysregulations in various physiological systems on the other hand [including the endocrine, immune, and autonomic nervous systems (Joynt et al., 2003; Lett et al., 2004)]. With respect to autonomic nervous system (ANS) dysregulation there is evidence to suggest that depressive symptoms are related to higher sympathetic nervous system (SNS) activation (Joynt et al., 2003), attenuated vagal tone (Carney et al., 2001; Hughes and Stoney, 2001; Rottenberg, 2007; Schwerdtfeger and Friedrich-Mai, 2009; Udupa et al., 2007), attenuated baroreceptorreflex-sensitivity (Watkins and Grossman, 1999), and elevated blood pressure (e.g., Grewen et al., 2004; Hamer et al., 2007; Light et al., 1998). These various effects may impose elevated load on the cardiovascular system, thus fostering the development of CVD. Of note, depressive symptoms have also been related to cardiovascular reactivity (CVR). According to the reactivity hypothesis,

∗ Corresponding author at: Department of Psychology, Karl-Franzens-University Graz, A-8010 Graz, Austria. Tel.: +43 3163804953; fax: +43 3163809807. E-mail address: [email protected] (A. Schwerdtfeger). 0301-0511/$ – see front matter © 2011 Elsevier B.V. All rights reserved. doi:10.1016/j.biopsycho.2011.05.009

repeated stress-related increases in cardiovascular function are assumed to accelerate a wear and tear on the artery walls, leading to endothelial dysfunction and, ultimately, CVD (Chida and Steptoe, 2010; Harris and Matthews, 2004; Schwartz et al., 2003). Hence, it seems reasonable to assume that elevated CVR constitutes one path through which depression could affect cardiovascular health. A considerable number of studies has been devoted to the relationship between depressive symptoms and CVR. Whereas some studies could observe elevated CVR to laboratory stressors in depressed individuals (e.g., Light et al., 1998; Matthews et al., 2005), other studies found that this effect was dependent on other psychological variables (e.g., aggression; Betensky and Contrada, 2010) or even failed to support this relationship (Taylor et al., 2006). Of note, a meta-analysis of 11 studies published until 2001 (Kibler and Ma, 2004) report positive, however, not reliable associations between depression and CVR, thus providing only limited support for the assumption that CVR links depression with adverse health outcomes. On the contrary, an increasing number of recently published studies found evidence for attenuated – and not elevated – CVR with increasing depression scores (e.g., Carroll et al., 2007; Phillips, 2011; Salomon et al., 2009; York et al., 2007). For example, using a mental arithmetic task as a laboratory stressor Carroll et al. (2007) found in a population study that individuals with elevated depressive symptoms showed lower systolic blood pressure (SBP) and heart rate (HR) responses. Essentially the same finding was reported by Phillips (2011). Similarly, York et al. (2007) could observe that depressed individuals with coronary artery disease exhibited smaller increases in HR and SBP during a public speaking task than their counterparts with comparably low depression

A. Schwerdtfeger, A.-K. Rosenkaimer / Biological Psychology 87 (2011) 430–438

scores. Hence, it appears that depression might be associated with diminished CVR during certain aversive encounters. However, the role of stressor type in studies on depression and CVR is not well understood. Generally, two types of laboratory stressors can be distinguished, namely active stressors (e.g., public speaking tasks, mental arithmetics) and passive stressors [e.g., the cold pressor task (CP) or mirror tracing; Hurwitz et al., 1993; Obrist, 1981]. Whereas active stressors are associated with a cardiovascular response pattern, which can be characterized by blood pressure increase, elevated cardiac contractility and cardiac output, as well as decreased peripheral resistance (indicating a beta-adrenergic response profile), the blood pressure increase to passive stressors is accompanied by attenuated cardiac output, and elevated peripheral resistance (indicating an alpha-adrenergic response profile; Brownley et al., 2000; Hurwitz et al., 1993). It is interesting to note here that attenuated CVR in depressed individuals was mainly found when beta-adrenergic stressors were applied (Carroll et al., 2007; Phillips, 2011; York et al., 2007), but there is little support for this finding for alpha-adrenergic stressors. In particular, Salomon et al. (2009) examined CVR to a public speaking task and a mirror tracing task in individuals diagnosed with major depressive disorder and healthy controls, thus allowing to contrast the effects of stressor type within the same study. In line with recent evidence, depressed individuals showed significantly lower SBP, HR, and cardiac output during the speech stressor, whereas the evidence for attenuated CVR was less clear for the mirror tracing task, which is an alpha-adrenergic stressor. Hence, blunted CVR in depressed individuals seems to depend on the type of stressor. Importantly, the finding of blunted CVR to active, betaadrenergic stressors in depressed individuals is entirely consistent with the phenomenon of a motivational deficit in depression. Depressed individuals show a deficit in approach-related behavior. For example, McFarland and Klein (2009) recently found that depressed individuals exhibited attenuated emotional reactivity to anticipated monetary rewards, but did not differ from non-depressed when they anticipated non-reward or punishment. In line with this evidence, Brinkmann and Gendolla (2008) could observe that depressive symptoms among otherwise healthy participants were associated with attenuated SBP reactivity in response to a difficult stress task but not in response to an easy task. The authors argued that individuals mobilize resources as long as success is possible and worthwhile. In the case of depression, negative mood functions as information for high task demand, resulting in effort deterioration and, correspondingly, lower SBP reactivity. Taken together, depressed individuals show an appetitive deficit in laboratory tasks and, consequently, may not invest much effort during active tasks, resulting in lower CVR. 1.1. Aim of the study The aim of this study was to examine CVR in non-clinical individuals with varying depression scores to different laboratory stressors. In order to investigate the specificity of the findings with respect to stressor type in more detail, participants were faced with both active and passive aversive encounters. We implemented three different stressors. There was an active beta-adrenergic stressor (public speaking task), in which participants were instructed to prepare and deliver a speech within a social-evaluative context (similar to the public speaking task used by Salomon et al., 2009), and two different alpha-adrenergic passive stressors (a CP task and a video viewing task). We decided to implement two different passive stressors for the following reasons: first, previous studies largely neglected passive stressors to provoke CVR in non-clinical individuals with depressive symptoms. Hence, there is a need for research applying different passive stressors to gain a broader view of blunted CVR in mildly depressed individuals. Second, passive

431

stressors usually are of little self-relevance, whereas active stressors are much more relevant to the self (e.g., via the evaluation of personal performance). Hence, when contrasting the role of stressor type to examine CVR there is a risk of confounding stressor type with self-relevance. Of note, personal relevance and negative self-views are a central feature of many theories of depression (e.g., Beck et al., 1979; Wisco, 2009). Thus, self-relevance could be more crucial for CVR as related to depressive symptoms than stressor type. Taken together, we were interested to examine CVR to two passive tasks which differed with respect to self-relevance. The CP was chosen as a physically challenging task with little self-relevance. In order to contrast the CP with a passive self-relevant stressor, we additionally implemented a video viewing task in which participants were asked to watch the videotape of them presenting the speech. Thus, the video viewing task mirrored the speech task, but this time it was a passive task, requiring no effort allocation. In line with previous evidence we expected that depressive symptoms would be associated with blunted CVR to the active task (i.e., less effort allocation and approach behavior in depressives when active task performance is required), but not during the passive non self-relevant task (e.g., Salomon et al., 2009). Moreover, if selfrelevance is the key to diminished CVR in depressive individuals, we would predict that depressive symptoms will also be related with attenuated reactivity to the viewing of the speech video, but not to the CP. Importantly, besides the well-studied cardiovascular system we also opted for recording electrodermal reactivity (EDR). Of note, electrodermal hyporeactivity in depressed individuals has been reported in a number of previous studies (e.g., Dawson et al., 1977; Donat and McCullough, 1983; Greenfield et al., 1963; Iacono et al., 1983; Lader and Wing, 1969; McCarron, 1973; Noble and Lader, 1971; Thorell, 2009; Zuckerman et al., 1968). Similar to blood pressure, electrodermal activity is predominately influenced by sympathetic nerve fibers (e.g., Boucsein, 1992). Hence, these findings might be interpreted in terms of a more generalized sympathetic nervous system dysfunction in individuals with depressive symptoms. Finally, we aimed to exploratively analyze physiological recovery. Of note, recent research suggests that especially cardiovascular recovery may be more important for physical health than peak reactivity to a challenge (Brosschot et al., 2006; Steptoe and Marmot, 2005). With respect to depression, the so-called perseverative cognition hypothesis (e.g., Brosschot, 2010; Brosschot et al., 2006) suggests that ruminating thoughts and worry may lead to sustained physiological activation, ultimately imposing risk for disease. However, recent research on diminished CVR in depression largely neglected recovery (e.g., Carroll et al., 2007; Phillips, 2011; York et al., 2007) or found no consistent association (Salomon et al., 2009), thus necessitating further research. 2. Methods 2.1. Participants Fifty-five volunteers (34 females) participated in the study. They had a mean age of 22.95 years (SD = 3.83) and a mean BMI of 22.55 (SD = 3.79). Twenty-nine percent of the sample were smokers and 69% reported regular physical activity. All participants were free of cardioactive and antidepressive medications. They were not allowed to consume caffeine or cigarettes 2 h prior to the experiment. Participants were recruited through advertisements at the university campus and oral communication. They received course credit for participating. 2.2. Stress tasks Three stress tasks were implemented to examine psychological and physiological reactivity. Throughout the stressor phases, the experimenter accompanied the participant in the room to provide ratings of affect and task performance. This procedure was implemented to enhance social-evaluative cues during task perfor-

432

A. Schwerdtfeger, A.-K. Rosenkaimer / Biological Psychology 87 (2011) 430–438

mance (e.g., Salomon et al., 2009). The following tasks were presented in serial order because a randomized presentation was not possible (the speech has to be recorded first before it can be presented on the video screen). 2.2.1. Public speaking As an active self-threatening stressor we used a public speaking task. This beta-adrenergic stressor has been shown to elicit relatively high, stable, and homogeneous reactivity (Al’Absi et al., 1997). Participants were instructed to prepare and deliver a public speech in front of a video camera. In particular, they were asked to elaborate on the pros and cons of a controversial social issue (medicide). To enhance social-evaluative cues, the experimenter accompanied the participant in the experimental room and evaluated emotional responses using paper and pencil. Task duration was 3 min. 2.2.2. Cold pressor A CP task was applied as a passive non-self-relevant stressor. The CP is perceived as painful and has been shown to increase cardiovascular responses to a considerable degree (e.g., Lovallo, 1975). The CP is a physical stressor and is known to provoke a consistent alpha-adrenergic response pattern (e.g., Saab et al., 1993). Participants were requested to immerse their dominant hand for 1 min in a bucket of ice water (4 ◦ C). The experimenter was present in the room and evaluated emotional responses. 2.2.3. Video viewing Participants were instructed to view the video of their speech performance on the computer screen. Therefore, they were shown the videotape of their speech from task 1 for 3 min. Passively viewing one’s own performance on video has been shown to provoke an alpha-adrenergic response pattern with comparably strong vascular reactivity (Hartley et al., 1999). To enhance social-evaluative cues, the experimenter accompanied the participant in the room and evaluated emotional responses. The video viewing task served as a passive self-relevant (i.e., psychological) stressor. The stressor tasks were embedded between baseline periods (3 min), preparation periods (3 min for the speech task and the video viewing task, 2 min for to the CP task), and recovery periods (3 min following each stressor). During each baseline period, participants viewed different episodes of an aquatic video to prevent carry-over effects (e.g., Piferi et al., 2000). Across all phases of the tasks, HR, SBP, DBP, and electrodermal readings were continuously recorded. 2.3. Measures 2.3.1. Depressive symptoms Depressive symptoms were assessed by means of the BDI (Beck and Steer, 1987). The BDI is a well validated questionnaire assessing cognitive, affective, and somatic symptoms of depression (Beck et al., 1988). We used a simplified German version of this scale, the BDI-V (Schmitt and Maes, 2000). It comprises 20 items with a 6-point frequency rating (0 = never, 5 = almost always). The mean score was 29.69 (SD = 17.25), which was significantly higher as compared to a normative German sample of N = 4494 participants [M = 20.4, SD = 14.2, t(54) = 3.99, p < .001; Schmitt et al., 2006]. On the basis of sensitivity and specificity estimates, Schmitt et al. (2006) proposed that a BDI-V score of 35 and above should be considered clinically critical. Applying this criterion in our sample resulted in approximately 26% of individuals with elevated depressive symptoms scores. The reliability of this scale for the present sample was good and in accordance with previous studies (Cronbach’s alpha = .93). 2.3.2. CVR and EDR Cardiovascular variables were recorded throughout the phases of the three tasks. We focused on HR, SBP, and DBP, since these are the main variables that have been used to assess CVR among depressed or dysphoric participants (e.g., Carroll et al., 2007; Phillips, 2011; Salomon et al., 2009). Blood pressure and HR were recorded by means of a tonometric device that allows continuous monitoring of beat-to-beat changes in blood pressure (Colin CBM-7000® ). Recording is based on arterial tonometry, which makes use of 30 cutaneously applied pressure sensors (piezoresistive transducers) that compress the top 15% of the underlying radial artery against the bone. From the resulting contact stress, the intralumenal blood pressure can be approximated (Drzewiecki et al., 1983). The sensor is placed with a wrist brace over the radial artery. Oscillographic cuff measurement is casually needed for calibration of the arterial tonometer. A normal-sized adult cuff was applied at the same arm site as the tonometric sensor. Advanced multiplexing electronics combined with automatic sensor positioning relative to the artery ensure optimal placement of the sensor. Two specific values are crucial for obtaining reliable data: hold-down pressure (HDP) and signal strength (SS). Whereas HDP defines the pressure applied to the artery in mmHg and should not exceed 140 mmHg or drop below 40 mmHg, SS refers to the strength of the tonometric signal in relation to the oscillographic signal. Maximum signal strength is 100%, and values should not drop below 60%. We calibrated the signal on 3 occasions: at the beginning of the experiment and following each recovery period. The mean HDP throughout the calibration periods varied between 108.15 and 112.07, and the mean SS varied between 76% and 83%,

thus indicating reliable recording. The Colin CBM-7000® has been proven to satisfy the standards of the Association for the Advancement of Medical Instrumentation (AAMI) for mean systolic and diastolic blood pressure measurements as well as the FDA (U.S. Food and Drug Administration) standard for intensive care units (Zorn et al., 1997). This device has also been validated in cardiovascular stress research where it has shown low artifact ratings and high accuracy (Nelesen and Dimsdale, 2002). Moreover, EDR was recorded by means of a Coulbourn® skin conductance coupler (S71-22), providing 0.5 V constant voltage. The number of non-specific skin conductance responses (NSSCRs) was recorded in AC mode with a time constant of 10 s. Hellige Ag/AgCl-electrodes (1.0 cm2 ) were placed at the non-dominant hand thenar and hypothenar by using adhesive collars. Electrodes were filled with EDAcream of 0.5% NaCl (TDE-246). The signal was sampled at 20 Hz and stored on a PC for further offline-analysis. 2.4. Procedure After arriving at the lab, informed consent was obtained and participants were told that they could leave if they wished to discontinue at any time, without giving a reason. Participants were then requested to wash their hands with water only and to dry them with hot air to assure comparability of skin conductance recordings across participants. Then the electrodes and blood pressure sensor were attached and thereafter questionnaires on demographic and lifestyle variables and the BDI were filled out. The experiment started with a blood pressure calibration, and thereafter the baseline period of 3 min was initialized during which participants viewed a calming aquatic video. The experiment proceeded with the instructions for the public speaking task, which was followed by a preparation period (3 min). Then, the camera was adjusted and the experimenter accompanied the participant in the lab. Participants presented the speech for 3 min. Then the recovery period followed. Before the next task began, a recalibration of the blood pressure recording was initialized, which was followed by another baseline period. Participants were then instructed to immerse their dominant hand in ice water for 1 min, but prior to this there was a preparation period of 2 min. After the preparation period, the experimenter entered the room and the task was started. Then another recovery period (3 min) followed. The next recalibration of the blood pressure recording was initialized, which was followed by a further baseline period (3 min). Thereafter, the next instructions appeared on the computer screen, introducing the upcoming video. Then a preparation period of 3 min followed. After that, the experimenter entered the room and the video was displayed. Finally, the recovery period was initialized (3 min) and thereafter the electrodes were detached and participants were thanked and debriefed. Participants were instructed not to communicate with other to-be participants about the content of this experiment. 2.5. Data recording and analysis Blood pressure and HR were averaged separately across each phase of the experiment, and NSSCRs were summed offline across the phases of the tasks. Responses were scored if amplitudes exceeded 0.02 ␮S (e.g., Boucsein, 1992; Schwerdtfeger, 2004). NSSCR was then aggregated across 1 min to yield a measure of NSSCR/min. Hypertensive participants (individuals consistently exceeding 145 mmHg systolic or 90 mmHg diastolic blood pressure throughout the calibration periods; N = 2) were excluded from the analyses of the cardiovascular variables (e.g., Harbison et al., 2009). Furthermore, 1 participant was excluded from the analyses of blood pressure because of extreme outliers (>4 SDs) in reactivity, leaving a total of 52 participants. Of note, 6 individuals had to be excluded from the analysis of blood pressure for the CP task because they showed sensor positioning shifts together with sudden drops in blood pressure, indicating gross postural changes during the immersion of their hand in ice water. We also monitored performance during the CP task and could observe that all but three participants endured the task for the entire 1-min period. However, three individuals (all females with BDI scores ranging from 22 to 41) withdrew their hand too early from the bucket of ice water (two participants withdrew after 30 s and one participant after 46 s). Importantly, excluding these participants from the analysis of CP reactivity did virtually not change any of the results presented in Table 3. Thus, we decided to include these participants in the analyses to not loose statistical power. LME models were calculated for HR, SBP, DBP, and NSSCR separately. For each physiological variable we analyzed each task separately and allowed heteroscedasticity with respect to task period for each participant (random intercepts and random slopes). Demographic variables (sex, age, BMI), depression (BDI), task period (0 = baseline/reference, 1 = preparation, 2 = speech/CP/video, 3 = recovery), and the interaction of depressive symptoms and task period were entered as fixed effects predictors, and participants were entered as random effects. LME models have been recommended in psychophysiological research (Bagiella et al., 2000), because they offer several advantages over traditional ANOVA-based algorithms (e.g., more efficient handling of missing data, more powerful tests that allow for modeling the error variance, including higher flexibility). We applied the open source language and environment for statistical computing R (version 2.9.0; R Development Core Team, 2009), using the lmer program of the lme4 package (version 0.99375-31; for an overview of lme4, see Bates, 2005). We report regression coefficients (b; absolute effect size), standard errors (SE), and their ratio (t statistic).

Note: BL = baseline, Prep. = preparation period, Speech = speech stressor, Rec. = recovery, CP = cold pressor, Video = video stressor.

Rec. Video

75.51 (11.64) 125.59 (24.93) 88.74 (18.16) 10.94 (4.52) 76.61 (11.95) 119.97 (20.21) 83.60 (15.08) 7.46 (4.41)

Prep. BL

72.80 (11.11) 113.26 (14.47) 79.98 (10.92) 4.68 (3.87) 73.14 (11.19) 116.95 (20.65) 82.52 (16.83) 3.09 (3.37)

Rec. CP

84.26 (14.58) 133.78 (30.29) 93.39 (23.83) 9.48 (5.47) 75.48 (12.40) 119.91 (18.25) 83.88 (14.14) 6.33 (3.84)

Prep. BL

74.08 (11.97) 114.54 (13.34) 81.10 (10.38) 4.69 (3.80) 75.08 (13.01) 114.19 (23.10) 74.91 (22.73) 4.91 (3.99)

Rec. Speech Prep. BL

Table 1 Means and standard deviations (in brackets) of the main variables throughout the experiment.

For the speech task we found significant increases from baseline to preparation and speech delivery for each cardiovascular variable with a subsequent decrease during recovery, suggesting that the stressor was effective. Please note that DBP during recovery was lower than during baseline. Moreover, the interaction of BDI and preparation was significant for HR, SBP, and DBP, documenting negative associations between depressive symptoms and CVR during preparation. The interaction of BDI and speech delivery was also significant for SBP and DBP, but failed to reach significance for HR. To examine these interaction effects further, we calculated simple-slope analyses. Therefore, we rescaled the BDI at the standard deviation, thus allowing us to analyze individuals high (1 SD above the mean) and low (1 SD below the mean) on the BDI, thereby controlling for all other covariates. Specifically, two continuous variables were calculated that were scaled to zero at either 1 SD above (i.e., BDI+) or 1 SD below the mean (i.e., BDI−). Then, two additional analyses were run in which the newly computed high and low BDI variables were separately entered into the equation replacing the original BDI variable. Importantly, this kind of analysis makes use of the whole sample size, thus retaining the same statistical power as the previous models. These analyses suggested that individuals high in depressive symptoms showed approximately 9 mmHG (preparation period) and 17 mmHg (speech) lower SBP reactivity than individuals low in depression. The difference in DBP was 5 mmHG for the preparation period and 10 mmHG for the speech. Further, individuals high in depressive symptoms showed a 5 BPM lower HR during preparation than individuals with comparably few depressive symptoms. For the CP task (Table 3) there were significant increases to both preparation and task performance for each cardiovascular variable. Moreover, CVR returned to baseline levels for SBP and DBP, and decreased below baseline level for HR. Of note, there were no significant interactions of BDI and any of the periods of the CP task (all |T|s < 1.55). Further, there were significant sex effects for blood pressure, indicating 9 mmHg lower SBP and 5 mmHg lower DBP in women as compared to men. Finally, during baseline both SBP and DBP were positively associated with BMI, indicating elevated blood pressure with higher BMI. For the video viewing task (Table 4) there were significant cardiovascular increases to preparation and task performance. Moreover, HR and DBP significantly decreased to baseline levels during recovery, whereas SBP exceeded baseline level during recovery. Importantly, there were significant interactions of BDI and preparation, BDI and video viewing, and BDI and recovery for SBP and DBP, respectively. Again, simple-slope analyses as described above were calculated to examine these interactions in more detail. We found that elevated BDI scores were accompanied by lower SBP and DBP reactivity throughout different phases

99.91 (16.83) 138.35 (27.62) 94.32 (21.07) 12.67 (3.53)

3.1. CVR

88.06 (16.33) 131.16 (20.89) 88.36 (18.49) 10.29 (4.92)

Table 1 shows the descriptive statistics of the physiological variables in the course of this experiment. The mixed-effects analyses of the various physiological reactivity measures are presented in Table 2 (public speaking task), Table 3 (CP), and Table 4 (video viewing task).

75.56 (12.74) 115.96 (16.48) 80.26 (15.18) 5.65 (4.53)

3. Results

HR SBP DBP NSSCR

As the formulas for the degrees of freedom for inferences based on t or F distributions do not apply in mixed-effects models, the calculation of p-values gets problematic and can lead to anti-conservative decisions. However, for comparably large data sets it has been recommended to interpret t values exceeding 2 SE as significant (e.g., Kliegl, 2007).

433

73.00 (11.16) 116.92 (20.50) 82.70 (17.60) 3.80 (3.46)

A. Schwerdtfeger, A.-K. Rosenkaimer / Biological Psychology 87 (2011) 430–438

434

A. Schwerdtfeger, A.-K. Rosenkaimer / Biological Psychology 87 (2011) 430–438

Table 2 Linear mixed-effects models for predicting heart rate (HR), systolic blood pressure (SBP), diastolic blood pressure (DBP), and the number of skin conductance responses (NSSCR; right side) to the speech task. Variable

HR ba

Intercept Sex (0 = male, 1 = female) Age BMI BDI Preparation (vs. BL) Speech (vs. BL) Recovery (vs. BL) BDI × preparation (vs. BL) BDI × speech (vs. BL) BDI × recovery (vs. BL)

73.83 2.44 0.64 −0.33 −0.02 12.42 23.80 −0.44 −0.15 −0.15 0.01

SBP SE 2.92 3.69 0.55 0.55 0.11 1.27 1.60 0.64 0.07 0.09 0.04

ba

T *

25.30 0.66 1.17 −0.59 −0.20 9.76* 14.92* −0.69 −2.02* −1.61 0.21

120.59 −7.20 −0.51 1.06 −0.07 15.34 23.32 −1.80 −0.27 −0.49 −0.21

DBP SE 3.61 4.62 0.69 0.69 0.13 1.65 2.76 2.67 0.09 0.16 0.15

ba

T *

33.40 −1.56 −0.74 1.53 −0.50 9.31* 8.46* −0.67 −2.86* −3.06* −1.37

80.24 0.30 −0.74 0.71 −0.18 8.02 14.44 −5.58 −0.15 −0.28 −0.15

NSSCR/min SE 3.45 4.42 0.66 0.66 0.13 1.29 1.96 2.31 0.07 0.11 0.13

T *

23.23 0.07 −1.13 1.07 −1.46 6.22* 7.39* −2.42* −2.02* −2.51* −1.15

ba

SE

T

5.69 0.15 −0.16 0.13 −0.06 4.60 6.90 −0.77 −0.05 −0.01 −0.004

0.84 0.93 0.14 0.14 0.04 0.38 0.49 0.33 0.02 0.03 0.02

6.74* 0.16 −1.13 0.90 −1.64 12.16* 14.10* −2.33* −2.35* −0.51 −0.22

Note: BDI = Beck Depression Inventory, BL = baseline. a Unstandardized partial regression coefficients. * p < .05. Table 3 Linear mixed-effects models for predicting heart rate (HR), systolic blood pressure (SBP), diastolic blood pressure (DBP), and the number of skin conductance responses (NSSCR; right side) to the cold pressor (CP). Variable

HR ba

Intercept Sex (0 = male, 1 = female) Age BMI BDI Preparation (vs. BL) CP (vs. BL) Recovery (vs. BL) BDI × preparation (vs. BL) BDI × CP (vs. BL) BDI × recovery (vs. BL)

73.90 −0.11 0.65 −0.23 0.01 1.45 10.24 −1.02 −0.006 0.002 0.01

SBP SE 2.68 3.33 0.50 0.50 0.10 0.37 1.24 0.42 0.02 0.07 0.02

ba

T *

27.59 −0.03 1.30 −0.46 0.11 3.93* 8.28* −2.40* −0.29 0.02 0.54

120.63 −9.62 0.11 1.33 0.10 5.39 18.98 2.90 0.04 −0.29 −0.08

DBP SE 2.30 2.84 0.42 0.43 0.09 1.11 3.33 1.64 0.06 0.19 0.09

ba

T *

52.37 −3.39* 0.27 3.11* 1.17 4.86* 5.70* 1.77 0.57 −1.54 −0.80

84.24 −5.03 0.06 1.16 0.06 2.78 12.95 1.74 0.02 −0.20 −0.06

NSSCR/min SE 1.89 2.28 0.34 0.34 0.07 0.84 2.78 1.41 0.05 0.16 0.08

T *

44.60 −2.21* 0.19 3.39* 0.79 3.30* 4.65* 1.23 0.40 −1.26 −0.76

ba

SE

T

5.32 −0.88 −0.04 0.005 −0.04 1.60 4.73 −1.65 0.005 −0.05 0.03

0.79 0.96 0.14 0.14 0.03 0.37 0.62 0.30 0.02 0.04 0.02

6.71* −0.91 −0.25 0.03 −1.16 4.33* 7.59* −5.56* 0.21 −1.49 1.68

Note: BDI = Beck Depression Inventory, BL = baseline. a Unstandardized partial regression coefficients. * p < .05.

of the task. In particular, holding all other variables constant SBP reactivity in individuals with elevated BDI scores (1 SD above the mean) was 6 mmHG (preparation), 11 mmHG (video viewing), and 8 mmHG (recovery) lower as compared to individuals with low BDI scores (1 SD below the mean). For DBP the respective differences were 4 mmHg (preparation), 7 mmHg (video viewing), and 5 mmHg (recovery). The SBP and DBP response curves for high and low depressive individuals throughout the tasks are depicted in Fig. 1.

3.2. EDR As revealed in Table 2, analyses of NSSCR for the speech task revealed significant increases from baseline to preparation and speech delivery, and a subsequent drop below baseline level during recovery. Further, there was a significant interaction of BDI and preparation, indicating lower EDR with increasing BDI scores. Again, simple-slope analyses were calculated to elucidate this effect. Individuals with comparably low BDI scores (1 SD above the

Table 4 Linear mixed-effects models for predicting heart rate (HR), systolic blood pressure (SBP), diastolic blood pressure (DBP), and the number of skin conductance responses (NSSCR; right side) to the viewing of the speech video. Variable

Intercept Sex (0 = male, 1 = female) Age BMI BDI Preparation (vs. BL) Video (vs. BL) Recovery (vs. BL) BDI × preparation (vs. BL) BDI × video (vs. BL) BDI × recovery (vs. BL)

HR

SBP

DBP

NSSCR/min

ba

SE

T

ba

SE

T

ba

SE

T

ba

SE

T

70.98 2.70 0.92 −0.37 −0.01 3.90 2.52 −0.11 −0.05 −0.03 0.003

2.51 3.16 0.47 0.47 0.09 0.61 0.73 0.40 0.04 0.04 0.02

28.27* 0.86 1.95 −0.79 −0.06 6.40* 3.47* −0.28 −1.52 −0.72 0.15

118.04 −7.98 0.81 0.32 −0.04 7.29 13.09 3.54 −0.17 −0.31 −0.24

2.99 3.78 0.56 0.57 0.11 1.33 2.32 1.72 0.08 0.13 0.10

39.52* −2.11* 1.44 0.56 −0.40 5.49* 5.64* 2.06* −2.28* −2.35* −2.38*

81.01 −1.81 0.67 0.20 −0.11 3.96 8.96 2.38 −0.12 −0.20 −0.20

2.22 2.79 0.41 0.42 0.08 0.91 1.53 1.37 0.05 0.09 0.08

36.48* −0.65 1.62 0.48 −1.36 4.37* 5.85* 1.74 −2.31* −2.31* −2.49*

5.44 −1.21 −0.09 0.04 −0.009 2.77 6.26 −0.85 −0.06 −0.06 0.002

0.83 1.01 0.15 0.15 0.03 0.39 0.47 0.30 0.02 0.03 0.02

6.53* −1.20 −0.59 0.24 −0.28 7.05* 13.38* −2.81* −2.48* −2.20* 0.10

Note: BDI = Beck Depression Inventory, BL = baseline. a Unstandardized partial regression coefficients. * p < .05.

A. Schwerdtfeger, A.-K. Rosenkaimer / Biological Psychology 87 (2011) 430–438 160

Speech

Cold Pressor

Video Viewing

16

Speech

Cold Pressor

435

Video Viewing BDI + BDI -

14

150

BDI + BDI -

12

NSSCR/min

SBP mmHg

140

130

120

10 8 6 4

110

2 0

0 BL PREP SP REC

BL PREP CP REC

BL PREP SP REC

BL PREP VID REC

110

Speech

Cold Pressor

Video Viewing

105

BDI + BDI -

100 95

DBP mmHg

BL PREP CP REC

BL PREP VID REC

TASK PERIOD

TASK PERIOD

90

Fig. 2. Number of non-specific skin conductance responses (NSSCR) of high (1 SD above the mean of the BDI; BDI+) and low (1 SD below the mean of the BDI; BDI−) depressive participants throughout the different tasks. Values are derived from the mixed effect models and are adjusted for gender, age, and BMI. Reactivity from baseline in response to the speech and the video viewing task was significantly diminished in individuals with relatively more depressive symptoms. Note: BL = baseline, PREP = preparation, SP = speech task, CP = cold pressor task, VID = video viewing task.

85 80

it was 5.23 NSSCRs. The adjusted NSSCR response curves for high and low depressed individuals are depicted in Fig. 2.

75 70 65

4. Discussion

0 BL PREP SP REC

BL PREP CP REC

BL PREP VID REC

TASK PERIOD Fig. 1. Systolic blood pressure (SBP; upper figure) and diastolic blood pressure (DBP, lower figure) of high (1 SD above the mean of the BDI; BDI+) and low (1 SD below the mean of the BDI; BDI−) depressive participants throughout the different tasks. Values are derived from the mixed effect models and are adjusted for gender, age, and BMI. Reactivity from baseline in response to the speech and the video viewing task was significantly diminished in individuals with relatively more depressive symptoms. Note: BL = baseline, PREP = preparation, SP = speech task, CP = cold pressor task, VID = video viewing task.

mean) showed an excess of 1.8 NSSCRs above those with comparably high BDI scores (1 SD below the mean). The interactions for the speech delivery and recovery periods, respectively, did not reach significance. For the CP task (Table 3) we found significant increases from baseline to both preparation and CP. During recovery, NSSCR was significantly reduced as compared to baseline. No other effects reached significance. In particular, there were no significant interactions of BDI and any of the periods of the task (all |T|s < .1.69). Finally, the analysis of the video viewing task (Table 4) revealed significant NSSCR increases from baseline to preparation and video viewing, with a subsequent drop below baseline level during recovery. Again, there was a significant BDI × preparation interaction, indicating lower EDR to preparation in individuals with elevated BDI scores. Simple-slope analyses were calculated as described earlier. The respective predicted values were 3.74 NSSCRs for individuals scoring low on the BDI (1 SD below the mean) and 1.79 NSSCRs for those scoring high on the BDI (1 SD above the mean). Importantly, the interaction for the video viewing was also significant, thus mirroring the effect of the preparation period. Specifically, for individuals scoring low on the BDI the predicted value was 7.29 NSSCRs and for individuals scoring high on the BDI

The aim of this study was to examine the relationship between depressive symptoms and physiological stress reactivity to different aversive encounters. Specifically, we were interested to investigate whether blunted physiological reactivity in non-clinical individuals with depressive symptoms can be found during active and passive aversive encounters, and whether self-relevance of the task plays a significant role. We found evidence for blunted CVR and EDR to self-relevant stressors in individuals with comparably high scores on the BDI. Of note, these effects were evident irrespective of the task being active (public speaking) or passive (viewing of the speech video). Hence, our study corroborates and extends previous evidence suggesting attenuated physiological reactivity in depressed individuals to psychological tasks (e.g., Carroll et al., 2007; Salomon et al., 2009; York et al., 2007). Importantly, our study adds to a growing number of recent research showing attenuated – and not elevated – CVR in depressive individuals in response to active aversive encounters (e.g., Phillips, 2011; Salomon et al., 2009; York et al., 2007). Hence, it seems reasonable to assume that blunted CVR in depressed individuals to active stressors could reflect attenuated effort allocation, thus indicating a motivational deficit. Of particular relevance for this hypothesis, however, were the analyses for the passive selfrelevant task. Viewing one’s own performance on the speech video does not involve effort allocation since this task is a passive challenge, which is accompanied by a vascular response profile (Hartley et al., 1999). However, attenuated CVR in individuals with elevated depressive symptoms scores was also evident during the passive video viewing task, thus challenging the interpretation of lower effort mobilization in individuals with depressive symptoms. Specifically, depressive symptoms were reliably related to lower SBP and DBP reactivity during preparation, video viewing and, partially, recovery. Given the passive nature of this task and the general pattern throughout the periods of this task, it seems unlikely that differences in effort mobilization would have contributed to this effect. Rather, the findings suggest that the self-relevant nature of

436

A. Schwerdtfeger, A.-K. Rosenkaimer / Biological Psychology 87 (2011) 430–438

the tasks may have contributed to the blunted physiological reactivity in mildly depressed individuals. This finding represents a challenge to established theories on depression, which suggest that depressive individuals are more vulnerable to self-relevant information (Beck et al., 1979; Wisco, 2009). It might follow that self-relevant stimuli should provoke stronger affective reactions in depressives. For example, in a study conducted by Rottenberg et al. (2005) normative (e.g., standardized film clips) and idiographic stimuli (e.g., a video of the participant talking about emotional moments in her/his life) were presented. Currently depressed individuals showed a stronger subjective reactivity to happy and sad idiographic video stimuli than to normative stimuli, however, there was no response modulation with respect to autonomic reactivity (heart rate, finger pulse transit time, respiration rate, skin conductance). In sum, this study provides modest evidence for higher reactivity to self-relevant stimuli in depressed individuals. However, it should be noted that, unlike in our study, the authors implemented only passive tasks without enhancing social-evaluative stress. Hence, the findings are not directly comparable to our study. Moreover, recent theories suggest that depressed individuals may show emotion context insensitivity, which is associated with flattened – and not elevated – emotional reactivity (Bylsma et al., 2008). This theory is grounded on the assumption that depressed individuals exhibit emotional disengagement to inhibit dangerous or wasteful actions in situations in which a goal is perceived as unreachable (Nesse, 2000). This goal might imply good performance in an active psychological task or maintaining a positive self-image in a passive psychological task. Although this theory might prove useful to explain the current findings, it should be noted that it was developed to predict emotional reactivity in patients suffering from clinical depression and may not be applicable to mildly depressed individuals. Hence, further research on physiological reactivity in mildly depressed individuals to a variety of both active and passive stressors is certainly warranted. In line with this reasoning, it should be emphasized that we did not find evidence for attenuated CVR in individuals with elevated depressive symptoms for the CP task. Of note, the CP task is entirely different from the video viewing task. Whereas the viewing of the speech video is a psychological stressor, which is challenging with respect to the self, the CP is a physical stressor with little self-relevance. Importantly, the study conducted by Salomon et al. (2009) could also find no robust association between depression and attenuated CVR to a non self-relevant passive stressor (mirror tracing task). Thus, in light of this evidence we would suggest that hyporeactivity in individuals with depressive symptoms may be limited to both active and passive psychologically aversive situations, but may not be prevalent during non-self-relevant physical stressors. Importantly, for EDR we found a striking resemblance to CVR responses throughout the tasks. Specifically, there was evidence for attenuated EDR in depressed individuals during both self-relevant (i.e., psychological) tasks but no difference for the CP task. We want to emphasize that this result is largely consistent with previous studies documenting diminished EDR to different external stimuli in depressed individuals (e.g., Dawson et al., 1977; Donat and McCullough, 1983; Greenfield et al., 1963; Iacono et al., 1983; Lader and Wing, 1969; McCarron, 1973; Noble and Lader, 1971; Thorell, 2009; Zuckerman et al., 1968). Of note, attenuated EDR in severely depressed individuals has frequently been found when the same stimulus was repetitively presented (so-called habituation paradigms; for a review, see Thorell, 2009). This pattern of result has been discussed to reflect a less developed neural image of the properties of the stimulus, thus impacting the anticipation and prediction of external information (Thorell, 2009). In this respect it might be interesting to note that in our study electrodermal

hyporeactivity in depressed individuals was predominant in the video viewing task but less pronounced during the prior speaking task. Hence, it might be speculated that these findings could indicate an impaired neural representation of the self in individuals with elevated depressive symptoms scores. Certainly, more studies are needed to examine EDR and EDR-habituation to self-relevant material in depressed individuals to verify or falsify this hypothesis. Of note, no reliable associations were found for HR, except a lower HR during preparation of the speech in individuals with relatively more depressive symptoms. HR in contrast to both blood pressure and electrodermal activity is more strongly influenced by parasympathetic efference. Therefore, our results may be interpreted in terms of a dysfunction of the sympathetic branch of the ANS in depressed individuals. It should be noted, however, that there is also considerable evidence for attenuated parasympathetic (i.e., vagal) innervation of the heart in depressed individuals (e.g., Carney et al., 2001; Hughes and Stoney, 2001; Rottenberg, 2007), thus suggesting a more general ANS dysregulation in depressed individuals. Our study does not allow a more comprehensive analysis of these effects, because other more distinct measures of ANS activity were beyond the scope of this study. To accomplish this, we would recommend to record and analyze heart rate variability as a more direct measure of sympathetic–parasympathetic balance and preejection period as a more distinctive sympathetic indicator, for instance. This way, one would be better able to disentangle parasympathetic and sympathetic dysregulations with respect to depressive symptoms. Of note, we did not find evidence for impaired physiological recovery in individuals with elevated depressive symptoms scores as suggested by the perseverative cognition hypothesis (Brosschot et al., 2006). Specifically, the interaction terms of BDI and recovery indicated no reliable positive relationship between depressive symptoms and physiological recovery for neither task. On the contrary, individuals with elevated depressive symptom scores showed lower blood pressure readings (SBP and DBP) during the recovery of the video viewing task. Although our findings seem to suggest that perseverative cognition did not significantly impact physiological recovery in depressed individuals, the results should be interpreted in light of the generally mild depression scores in this sample and the rather short duration of the recovery period (3 min). In particular, it seems reasonable to assume that perseverative cognition is more prevalent in severely depressed individuals and that more extended recording periods following a psychological stressor would allow a more fine-grained analysis of this hypothesis. 4.1. Limitations and future directions Although our study provides considerable evidence for attenuated reactivity in various variables of the ANS to psychological stressors in individuals with depressive symptoms, these studies are not free of limitations. First, it should be noted that we used a self-report instrument to measure depression in this experiment, thus we could not verify if any of the participants would have met criteria for a diagnosis of depression. Hence, the generalizability of the present data to clinical depression is uncertain. However, the investigation of physiological reactivity in depression is challenged by the medication status of the participants. Specifically, antidepressive medication has been shown to significantly impact CVR (e.g., Hallas and Thornton, 2000). Hence, we decided to recruit a sample of non-clinical, medication-free participants with varying depressive symptoms scores. Nonetheless, the findings of Salomon et al. (2009), who analyzed a clinical sample are generally in accordance with our results. Notwithstanding this evidence, it would be useful to further examine physiological reactivity in more severely depressed individuals (including clinical depressed individuals) to verify the generalizability of the findings.

A. Schwerdtfeger, A.-K. Rosenkaimer / Biological Psychology 87 (2011) 430–438

Second, it should be mentioned that the sample size was moderate and there is a need for replication. In particular, some participants had to be excluded from the analysis of the CP task, because they showed postural changes leading to unreliable blood pressure recordings. It might be speculated whether a larger sample size might have resulted in significant differences between more and less depressed individuals also for the CP task. Hence, future studies should aim to replicate our findings with larger sample sizes. Third, it should be noted that the order of the tasks was not counterbalanced. Thus we cannot rule out that contrast effects or familiarization with the procedure in general could have affected the results to some degree. However, fully counterbalancing the order of the tasks was not feasible because the video viewing task was dependent on the recording of the speech. Finally, the present work could be extended by fully counterbalancing self-relevance and stressor type (active vs. passive). However, whereas it might be reasonable to implement self-relevant and non-self-relevant passive stressors, it may be challenging to introduce active tasks without any self-relevance. This could probably be achieved by manipulating social-evaluative cues to be present (like in our study) or completely absent. 4.2. Conclusion Despite this criticism, our study provides evidence for the hypothesis of attenuated SNS reactivity in individuals with depressive symptoms. In particular, blood pressure and skin conductance reactivity to both active and passive self-relevant (i.e., psychological) stressors were attenuated in individuals with comparably more depressive symptoms. Hence, this study contributes to a growing body of literature suggesting diminished ANS reactivity in depressives, which may also generalize to passive self-relevant encounters. Acknowledgement We would like to express our thanks to Boris Egloff for his valuable comments on an earlier draft of this article. References Al’Absi, M., Bongard, S., Buchanan, T., Pincomb, G.A., Licinio, J., Lovallo, W.R., 1997. Cardiovascular and neuroendocrine adjustment to public speaking and mental arithmetic stressors. Psychophysiology 34 (3), 266–275. Bagiella, E., Sloan, R.P., Heitjan, D.F., 2000. Mixed-effects models in psychophysiology. Psychophysiology 37 (1), 13–20. Barth, J., Schumacher, M., Herrmann-Lingen, C., 2004. Depression as a risk factor for mortality in patients with coronary heart disease: a meta-analysis. Psychosomatic Medicine 66 (6), 802–813. Bates, D.M., 2005. Fitting linear mixed models in R. R News 5, 27–30. Beck, A.T., Rush, A.J., Shaw, B.F., Emery, G., 1979. Cognitive Therapy of Depression. Guilford Press, New York. Beck, A.T., Steer, R.A., 1987. Beck Depression Inventory - Manual. San Antonio: the psychological association. Beck, A.T., Steer, R.A., Garbin, M.G., 1988. Psychometric properties of the Beck Depression Inventory: twenty-five years of evaluation. Clinical Psychology Review 8 (1), 77–100. Betensky, J.D., Contrada, R.J., 2010. Depressive symptoms, trait aggression, and cardiovascular reactivity to a laboratory stressor. Annals of Behavioral Medicine 39 (2), 184–191. Boucsein, W., 1992. Electrodermal Activity. Plenum Press, New York. Brinkmann, K., Gendolla, G.H.E., 2008. Does depression interfere with effort mobilization? Effects of dysphoria and task difficulty on cardiovascular response. Journal of Personality and Social Psychology 94 (1), 147–157. Brosschot, J.F., 2010. Markers of chronic stress: prolonged physiological activation and (un)conscious perseverative cognition. Neuroscience and Biobehavioral Reviews 35, 46–50. Brosschot, J.F., Gerin, W., Thayer, J.F., 2006. The perseverative cognition hypothesis: a review of worry, prolonged stress-related physiological activation, and health. Journal of Psychosomatic Research 60, 113–124. Brownley, K.A., Hurwitz, B.E., Schneiderman, N., 2000. Cardiovascular psychophysiology: function, methodology, and use in pathophysiological investigation. In:

437

Cacioppo, J.T., Tassinary, L.G., Berntson, G.G. (Eds.), Handbook of Psychophysiology. Cambridge University Press, New York, pp. 224–264. Bylsma, L.M., Morris, B.H., Rottenberg, J., 2008. A meta-analysis of emotional reactivity in major depressive disorder. Clinical Psychology Review 28 (4), 676–691. Carney, R.M., Blumenthal, J.A., Stein, P.K., Watkins, L., Catellier, D., Berkman, L.F., et al., 2001. Depression, heart rate variability, and acute myocardial infarction. Circulation 104 (17), 2024–2028. Carroll, D., Phillips, A.C., Der, G., 2007. Symptoms of depression and cardiovascular reactions to acute psychological stress: evidence from a population study. Biological Psychology 75 (1), 68–74. Chida, Y., Steptoe, A., 2010. Greater cardiovascular responses to laboratory mental stress are associated with poor subsequent cardiovascular risk status. A metaanalysis of prospective evidence. Hypertension 55 (4), 1026–1031. Dawson, M.E., Schell, A.M., Catania, J.J., 1977. Autonomic correlates of depression and clinical improvement following electroconvulsive shock therapy. Psychophysiology 14 (6), 569–578. Donat, D.C., McCullough, J.P., 1983. Psychophysiological discriminants of depression at rest and in response to stress. Journal of Clinical Psychology 39 (3), 315–320. Drzewiecki, G.M., Melbin, J., Noordergraaf, A., 1983. Arterial tonometry: review and analysis. Journal of Biomechanics 16 (2), 141–152. Greenfield, N.S., Katz, D., Alexander, A.A., Roessler, R., 1963. The relationship between physiological and psychological responsivity: depression and galvanic skin response. Journal of Nervous and Mental Disease 136 (6), 535–539. Grewen, K.M., Girdler, S.S., Hinderliter, A., Light, K.C., 2004. Depressive symptoms are related to higher ambulatory blood pressure in people with a family history of hypertension. Psychosomatic Medicine 66 (1), 9–16. Hallas, C.N., Thornton, E.W., 2000. Effects of SSRI antidepressant treatment upon mood status and blood pressure reactivity to demand in coronary patients. Coronary Health Care 4 (1), 2–9. Hamer, M., Tanaka, G., Okamura, H., Tsuda, A., Steptoe, A., 2007. The effects of depressive symptoms on cardiovascular and catecholamine responses to the induction of depressive mood. Biological Psychology 74 (1), 20–25. Harbison, J.A., Walsh, S., Kenny, R.A., 2009. Hypertension and daytime hypotension found on ambulatory blood pressure is associated with fatigue following stroke and TIA. QJM: An International Journal of Medicine 102 (2), 109–115. Harris, K.F., Matthews, K.A., 2004. Interactions between autonomic nervous system activity and endothelial function: a model for the development of cardiovascular disease. Psychosomatic Medicine 66 (2), 153–164. Hartley, T.R., Ginsburg, G.P., Heffner, K., 1999. Self presentation and cardiovascular reactivity. International Journal of Psychophysiology 32 (1), 75–88. Hughes, J.W., Stoney, C.M., 2001. Depressed mood is related to high-frequency heart rate variability during stressors. Psychosomatic Medicine 62 (6), 796–803. Hurwitz, B.E., Nelesen, R.A., Saab, P.G., et al., 1993. Differential patterns of dynamic cardiovascular regulation as a function of task. Biological Psychology 36 (1–2), 75–95. Iacono, W.G., Lykken, D., Peloquin, L., Lumry, A., Valentine, R.H., Tuason, V., 1983. Electrodermal activity in euthymic unipolar and bipolar affective disorders. A possible marker for depression. Archives of General Psychiatry 40 (5), 557–565. Joynt, K.E., Whellan, D.J., O’Connor, C.M., 2003. Depression and cardiovascular disease: mechanisms of interaction. Biological Psychiatry 54 (3), 248–261. Kibler, J.L., Ma, M., 2004. Depressive symptoms and cardiovascular reactivity to laboratory behavioral stress. International Journal of Behavioral Medicine 11 (2), 81–87. Kliegl, R., 2007. Towards a perceptual-span theory of distributed processing in reading: a reply to Rayner, Pollatsek, Drieghe, Slattery, & Reichle (2007). Journal of Experimental Psychology: General 136 (3), 530–537. Lader, M.H., Wing, L., 1969. Physiological measures in agitated and retarded depressed patients. Journal of Psychiatric Research 7 (2), 89–100. Lett, H.S., Blumenthal, J.A., Babyak, M.A., Sherwood, A., Strauman, T., Robins, C., et al., 2004. Depression as a risk factor for coronary artery disease: evidence, mechanisms, and treatment. Psychosomatic Medicine 66 (3), 305–315. Light, K.C., Kothandapani, R.V., Allen, M.T., 1998. Enhanced cardiovascular and catecholamine responses in women with depressive symptoms. International Journal of Psychophysiology 28 (2), 157–166. Lovallo, W., 1975. The cold pressor test and autonomic function: a review and integration. Psychophysiology 12 (3), 268–282. Matthews, S.C., Nelesen, R.A., Dimsdale, J.E., 2005. Depressive symptoms are associated with increased systemic vascular resistance to stress. Psychosomatic Medicine 67 (4), 509–513. McCarron, L.T., 1973. Psychophysiological discriminants of reactive depression. Psychophysiology 10 (3), 223–229. McFarland, B.R., Klein, D.N., 2009. Emotional reactivity in depression: diminished responsiveness to anticipated reward but not to anticipated punishment or to nonreward or avoidance. Depression and Anxiety 26 (2), 117–122. Nelesen, R.A., Dimsdale, J.E., 2002. Use of arterial tonometric continuous blood pressure measurement in cardiovascular reactivity studies. Blood Pressure Monitoring 7 (5), 259–263. Nesse, R.M., 2000. Is depression an adaptation? Archives of General Psychiatry 57 (1), 14–20. Noble, P., Lader, M., 1971. The symptomatic correlates of the skin conductance changes in depression. Journal of Psychiatric Research 9 (1), 61–69. Obrist, P.A., 1981. Cardiovascular Psychophysiology: A Perspective. Plenum Press, New York. Piferi, R.L., Kline, K.A., Younger, J., Lawler, K.A., 2000. An alternative approach for achieving cardiovascular baseline: viewing an aquatic video. International Journal of Psychophysiology 37 (2), 207–217.

438

A. Schwerdtfeger, A.-K. Rosenkaimer / Biological Psychology 87 (2011) 430–438

Phillips, A.C., 2011. Blunted cardiovascular reactivity related to depression, obesity, and self-reported health. Biological Psychology 86 (2), 106–113. R Development Core Team, 2009. R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing, Vienna, http://www.Rproject.org. Rottenberg, J., 2007. Cardiac vagal control in depression: a critical analysis. Biological Psychology 74 (2), 200–211. Rottenberg, J., Gross, J.J., Gotlib, I.H., 2005. Emotion context insensitivity in major depressive disorder. Journal of Abnormal Psychology 114 (4), 627–639. Rugulies, R., 2002. Depression as a predictor for coronary heart disease: a review and meta-analysis. American Journal of Preventive Medicine 23 (1), 51–61. Saab, P.G., Llabre, M.M., Hurwitz, B.E., Schneiderman, N., Wohlgemuth, W., Durel, L.A., Massie, C., Nagel, J., 1993. The cold pressor test: vascular and myocardial response patterns and their stability. Psychophysiology 30 (4), 366–373. Salomon, R., Clift, A., Karlsdottir, M., Rottenberg, J., 2009. Major depressive disorder is associated with attenuated cardiovascular reactivity and impaired recovery among those free of cardiovascular disease. Health Psychology 28 (2), 157–165. Schmitt, M., Altstötter-Gleich, C., Hinz, A., Maes, J., Brähler, E., 2006. Normwerte für das vereinfachte Beck-Depressions-Inventar (BDI-V) in der Allgemeinbevölkerung (Norms for the simplified Beck Depression Inventory (BDI-V) in a non-clinical population). Diagnostica 52 (2), 51–59. Schmitt, M., Maes, J., 2000. Vorschlag zur Vereinfachung des Beck-DepressionsInventars (BDI) (Suggestion for a simplification of the Beck Depression Inventory (BDI)). Diagnostica 46 (1), 38–46. Schwartz, A.R., Gerin, W., Davidson, K.W., Pickering, G.T., Brosschot, J.F., Thayer, J.F., et al., 2003. Toward a causal model of cardiovascular responses to stress and the development of cardiovascular disease. Psychosomatic Medicine 65 (1), 22–35. Schwerdtfeger, A., 2004. Predicting autonomic reactivity to public speaking: don’t get fixed on self-report data. International Journal of Psychophysiology 52 (3), 217–224. Schwerdtfeger, A., Friedrich-Mai, P., 2009. Social interaction moderates the relationship between depressive mood and heart rate variability: evidence from an ambulatory monitoring study. Health Psychology 28 (4), 501–509.

Steptoe, A., Marmot, M., 2005. Impaired cardiovascular recovery following stress predicts 3-year increases in blood pressure. Journal of Hypertension 23 (3), 529–536. Taylor, C.B., Conrad, A., Wilhelm, F.H., Neri, E., DeLorenzo, A., Kramer, M.A., et al., 2006. Psychophysiological and cortisol responses to psychological stress in depressed and nondepressed older men and women with elevated cardiovascular disease risk. Psychosomatic Medicine 68 (4), 538–546. Thorell, L.-H., 2009. Valid electrodermal hyporeactivity for depressive suicidal propensity offers links to cognitive theory. Acta Psychiatrica Scandinavica 119 (5), 338–349. Udupa, K., Sathyaprabha, T.N., Thirthalli, J., Kishore, K.R., Lavekar, G.S., Raju, T.R., Gangadhar, B.N., 2007. Alteration of cardiac autonomic functions in patients with major depression: a study using heart rate variability measures. Journal of Affective Disorders 100 (1–3), 137–141. Watkins, L.L., Grossman, P., 1999. Association of depressive symptoms with reduced baroreflex cardiac control in coronary artery disease. American Heart Journal 137 (3), 453–457. Wisco, B.E., 2009. Depressive cognition: self-reference and depth of processing. Clinical Psychology Review 29 (4), 382–392. Wulsin, L.R., Singal, B.M., 2003. Do depressive symptoms increase the risk for the onset of coronary disease? A systematic quantitative review. Psychosomatic Medicine 65 (2), 201–210. York, K.M., Hassan, M., Li, Q., Li, H., Fillingim, R.B., Sheps, D.S., 2007. Coronary artery disease and depression: patients with more depressive symptoms have lower cardiovascular reactivity during laboratory-induced mental stress. Psychosomatic Medicine 69 (6), 521–528. Zorn, E.A., Wilson, M.B., Angel, J.J., Zanella, J., Alpert, B.S., 1997. Validation of an automated arterial tonometry monitor using association for the advancement of medical instrumentation standards. Blood Pressure Monitoring 2 (4), 185–188. Zuckerman, M., Persky, H., Curtis, G.C., 1968. Relationships among anxiety, depression, hostility, and autonomic variables. Journal of Nervous and Mental Disease 146 (6), 481–487.