Psychophysiological Reactivity and PTSD Symptom Severity among Young Women

Abstract

Mixed findings with regard to cardiovascular reactivity (CVR) and posttraumatic stress disorder (PTSD) have suggested a need to further explore the nature of this relationship and factors that might explain differences in reactivity across and within samples. In the present study, the severity of PTSD symptoms was investigated in relation to CVR among young women. In addition, we examined whether severity within PTSD symptom clusters and level of dissociative symptoms were related to CVR. Heart rate, systolic and diastolic blood pressure, cardiac output (CO) and total peripheral resistance (TPR) reactivity in response to an oral speaking task were assessed for 58 young trauma-exposed civilian women with varying levels of PTSD symptomatology (from no symptoms to high severity of PTSD). The PTSD severity sub-scores for the DSM-V symptom clusters and total PTSD severity were based on structured interview (Clinician Administered PTSD Scale), and dissociative symptoms were assessed using the Dissociative Experiences Scale. Severity of total PTSD symptoms was associated with greater CO reactivity (r = .48, p < .01) and lower TPR reactivity (r = −.50, p < .01). Significant associations were not observed for heart rate or blood pressure. Results did not vary according to severity of symptoms within PTSD symptom cluster, with correlations for CO reactivity ranging from .40 to .49 and correlations for TPR reactivity ranging from −.40 to −.50 within symptom clusters. Dissociative symptoms were not significantly correlated with the CVR measures. Results partially supported the expectation that PTSD severity is one factor that would be associated with CVR, and suggest that reactivity for the underlying components of blood pressure (CO and TPR) provide additional information in probing stress reactivity in PTSD.

Introduction

Evidence in the literature continues to provide support for an association between exposure to traumatic events and cardiovascular disease (CVD; Boscarino, 2008b; Goodwin & Stein, 2004; Kubzansky, Koenen, Jones & Eaton, 2009; Spitzer et al., 2009; Violanti et al., 2006). Studies have also found support for higher odds for individuals with posttraumatic stress disorder (PTSD) to suffer CVD related mortality (OR = 2.4; Boscarino, 2008a), angina and heart failure (OR = 2.4 and OR = 3.4, respectively; Spitzer et al., 2009), functional limitations (OR = 2.2), greater symptom burden (OR = 1.9), and lower quality of life (OR = 2.5), than individuals without PTSD (Cohen et al., 2009). In particular, women with PTSD appear to be at increased risk for the development of CVD. Women have been reported to have double the lifetime prevalence rates of PTSD compared with men (Kessler et al., 2005). Seng, Clark, McCarthy and Ronis (2006) noted a dose-response association between physical comorbidity and the severity and chronicity of PTSD. Additionally, women with five or more PTSD symptoms have a threefold higher risk of CVD than women with no symptoms (Kubzansky et al., 2009). Despite this evidence, a meta-analytic review examining PTSD and its psychophysiological correlates revealed a paucity of studies with data for civilian women (Pole, 2007). These findings highlight the need to better understand the factors that contribute to the development of CVD in civilian women with PTSD.

Exaggerated cardiovascular reactivity (CVR) in individuals with PTSD has been identified as one of the potential mechanism responsible for the PTSD-CVD relationship (Buckley & Kaloupek, 2001; Hopper, Spinazzola, Simpson, & Van der Kolk, 2006; Hughes, Dennis, & Beckham, 2007). However, other studies have yielded mixed findings, demonstrating attenuated or blunted CVR responses (Ginty, Phillips, Roseboom, Carroll, & de Rooij, 2012; Phillips, Ginty, & Hughes, 2013). Historically, the inconsistencies found in CVR studies may have been related to the use of only blood pressure (systolic and diastolic) and heart rate (HR) measures (Larsen, Schneiderman, & Pasin, 1986; Manuck, 1994; Pickering & Gerin, 1990; Sherwood & Turner, 1995). Manuck, (1994) has long recommended that the underlying components of blood pressure, mainly cardiac output (CO) and total peripheral resistance (TPR), provide for a better measurement of CVR. Impedance cardiography (ICG) is recommended as a reliable, non-invasive method for assessing HR and the CO and TPR components of blood pressure (Saab, Llabre, Hurwitz, & Frame, 1992; Van De Water, Miller, Vogel, Mount, & Dalton, 2003; Yu, Nelesen, Ziegler & Dimsdale, 2001). Compared to more invasive measurement techniques, ICG allows for observation of circulation without disruption in the identification of these critical factors (Summers, Shoemaker, Peacock, Ander, & Coleman, 2003). Results from a study examining the association between mood states and hemodynamic variables noted that ICG-derived measurement of CO and TPR provided evidence for a subtle association with mood states, while HR, SBP and DBP were not able to do so (Yu et al., 2001). Recommendations in the literature support the use of more sophisticated measurement methods in order to develop a better understanding of the complex relationship between PTSD symptoms, CVR and CVD endpoints (Pole, 2007). Using ICG provides an efficient way to gather CO and TPR data in order to get a clearer picture of this relationship.

Research using the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition, Text Revision (DSM-IV-TR; APA, 2000), to make categorical determinations of PTSD diagnosis has not accounted for differences in severity of the symptom clusters within the PTSD diagnosis (hyperarousal, re-experiencing, and avoidance). The few studies that have incorporated symptom clusters have found differences in health perception, quality of life, physical health problems and symptoms, sleep disturbances, and CVR variables (SBP, DBP, and HR), related to differences in symptom cluster severity (Babson et al, 2011; Doctor, Zoellner, & Feeny, 2011; Kimerling, Clum, & Wolfe, 2000; Wolfe et al., 2000; Woods, Hall, Campbell, & Angott, 2008). Kimerling et al., (2000) found support for an association between greater hyperarousal symptoms and poorer health perception. Doctor et al. (2011) reported that only increased hyperarousal symptoms were associated with lower quality of life, while Zoellner et al. (2000), reported only the reexperiencing cluster was associated with self-reported physical symptoms. Woods et al., (2008) found evidence for a relationship that the avoidance cluster was predictive of physical health problems (i.e. sleep, stress, gynecological, and neuromuscular).

Lanius et al. (2010) proposed a theory of two subtypes for PTSD, a reexperiencing/hyperarousal subtype and a dissociative subtype. Further studies suggest that there are gender differences, with a predisposition for men towards a hyperarousal pattern, and a more dissociative pattern for women (Lanius et al., 2010; Olff, Langland, Net, & Gersons, 2007). Research examining associations of dissociative symptoms with physiological reactivity is also sparse, and with conflicting results (D’Andrea, Pole, DePierro, Freed & Wallace, 2013; Halligan, Michael, Wilhelm, Clark & Ehler, 2006; Sack, Cillien, & Hopper, 2012; Schalinski, Moran, Schauer & Elbert, 2014). Schalinski et al., (2014) examined the relationship between PTSD symptom severity, shutdown dissociation, and HR response, in a group of female physical and sexual assault survivors. They compared responses to unpleasant, neutral or pleasant stimuli, with results revealing a positive correlation between higher symptom severity and mean HR. However, no significant associations were found for shutdown dissociation and HR in the three conditions. In contrast, Halligan et al., (2006) found reduced HR responses in male and female survivors of physical and sexual assault with PTSD compared to survivors without PTSD; these results also indicated a significant negative relationship between dissociation and HR reactivity. A study of survivors with a variety of traumas compared participants exhibiting high reexperiencing and high acute dissociative symptoms to those with high reexperiencing and low dissociative symptoms, and found significantly lower HR reactivity in the high dissociation group (Sack et al., 2012). This line of research has led to changes in the criteria for PTSD in DSM-5 (APA, 2013; Friedman, Resick, Bryant, & Brewin, 2011; Miller et al., 2013), which further divided PTSD symptoms into four clusters (arousal and reactivity, intrusions, cognition and mood, and avoidance). Additionally, the specifier “with depersonalization and derealization symptoms”, has been added to identify individuals with the dissociative subtype of PTSD (APA, 2013). These studies highlight the need to further examine the relationship between symptoms of dissociation and PTSD symptom clusters.

The overall purpose of the current study was to provide a more comprehensive exploration of the factors involved in CVR reactivity among young women with PTSD (a population that has been under-studied with regard to CVR reactivity and CVD risk). We utilized impedance cardiography to explore HR, SBP and DBP as well as to form a clearer picture of the underlying components of BP (CO and TPR). Additionally, we examined the relationship of CVR with both dissociation and PTSD severity within each DSM-5 symptom cluster (re-experiencing, avoidance, negative cognitions and mood, arousal).

Method

Participants

Participants were 58 young (ages 19-40; M ± SD = 30 ± 8) civilian women recruited via posting of flyers, advertisement in scientific and public websites, and referrals from a community mental health clinic. Participants’ self-reported ethnicities were 50% Caucasian, 10% African American, 24% Hispanic White, 9% Caribbean Black, and 7% other. Participants were selected based on: 1) no history of major chronic illness, 2) pre-menopausal, 3) not currently taking any medication which might significantly influence the physiological measures (e.g. cholesterol lowering agents, beta-blockers, muscle relaxants), and 4) had the ability to read and speak English fluently. The study was open to participants with and without trauma histories. However, all participants reported exposure to at least one potentially traumatic event (see descriptive data section for information about PTSD symptoms/diagnosis).

Psychological Measures

Participants were invited for a clinical interview and assessment session. They completed a demographic questionnaire and were assessed using the Structured Clinical Interview for DSM-IV-TR Axis 1 Disorders (SCID; First & Gibbon, 2004) to determine axis 1 disorders aside from PTSD for inclusion/exclusion.

Prior to the structured interview for PTSD, participants completed an assessment of prior exposure to potentially traumatic events using the Trauma Life Event Questionnaire (TLEQ; Kubany et al., 2000). The TLEQ is a brief survey used to assess a broad range of potentially traumatic events such as motor vehicle accidents, combat, physical assault, and sexual abuse. The TLEQ has yielded good reliability, and good concurrent and predictive validity (Kubany et al., 2000). Posttraumatic stress disorder symptoms and severity were assessed using the Clinician-Administered PTSD Scale (CAPS), a structured interview that is an international standard in diagnostic assessments for PTSD (Blake et al., 1995). Only those with PTSD, unipolar depressive disorders, or no mental illness symptoms were selected for participation. Individuals with any other Axis I disorders were excluded.

The Dissociative Experiences Scale (DES) was administered to characterize dissociative symptoms. The DES is a 28-item rating scale with items scored on a scale of 0% to 100%, which indicate how often a symptom is experienced (van Ijzendoorn & Schuengel, 1996).

Physiological Assessment

Participants meeting inclusion criteria were invited for a second session measuring physiological responses to a speech delivery task. Impedance cardiography, automated BP measurement, and electrocardiogram (ECG) were used to asses CVR. The ECG electrodes were placed in a standard 3-lead configuration to identify the onset of electromechanical systole and to record HR. Impedance cardiography with standard full band tetrapolar configuration was used to records CVR measures. Four impedance cardiograph tape electrodes were placed at the neck (Z₁, Z₂) and the thorax (Z₃, Z₄) in order to assess CO, TPR and thoracic impedance. Z₁ was placed superior to the suprasternal notch of the thorax, Z₂ was placed exactly 3 cm above it, Z₃ was placed at the xiphoid process, and Z₄ was placed exactly 5 cm below it. The automated BP cuff (Tango; Suntech Med. Instr., Inc.) was attached to the dominant arm. Impedance cardiography, BP, and ECG recordings were analyzed for data scoring and reduction.

Physiological Data Sampling

The ECG, dz/dt and Z_o signals were collected continuously during the last four minutes of the pre-stressor baseline, and throughout the reactivity period. Mean arterial BP (MAP) is incorporated into the calculation of TPR, and BP measurement was initiated at the 11:00 min, 12:30 min and 14:00 min periods of the 15 min baseline, and 15 sec and 1:45 min point of the stressor period. The measures are initiated in this manner to allow ample time for cuff inflation and to appropriately reflect the physiological activity during the designated phase.

Physiological Data Reduction

The dz/dt signal and Z_o calibration signal were computer-stored during data collection for later conversion of the dz/dt to the unit of ohm/sec and derivation of stroke volume (ml). Ensemble averages of the cardiac cycles in the converted impedance cardiogram permitted the derivation of CO and the calculation of TPR as (MAP/CO) • 80 (dyne-s/cm⁵). For each physiological variable, a single average was calculated for the pre-task baseline and stressor period by averaging the last three baseline readings and the two stress task readings. Reactivity scores were derived by subtracting the respective baseline value from the stress task value.

Stress Induction

The delivery of a speech task was employed to induce psychological stress. Several forms of this task have been utilized to effectively elicit physiological arousal (Baggett, Saab, & Carver, 1996; Ramadan et al., 2013). The participants were instructed to formulate and deliver a speech in response to a hypothetical situation where a store security guard accuses them of stealing a belt that they are wearing (which they had purchased at the store three weeks prior), and to pretend that they requested to speak to the store manager and told him/her their side of the story. Participants were given 3 minutes to formulate the speech mentally, and were then asked to present the story aloud during a subsequent 3-minute period. The instructions indicated that performance would be videotaped and rated by experts on the basis of poise, speaking style, and appearance. The participant was given five elements around which to organize the story. The location of the camera was indicated prior to beginning the task. The participant was instructed when to begin the 3-min speech and when to stop. Standard prompts were provided by the experimenter to continue the speech, even if repetition of material was necessary, if the time had not elapsed. Upon completion of physiological measures, the recording equipment was removed and the experimenter explained the overall purpose of the study and answered any remaining questions the participant might have about the assessment. This speech delivery task has been used to produce both cardiac and vascular reactivity.

Results

Descriptive Data

The descriptive data for the primary study variables are depicted in Table 1. The PTSD symptom levels varied markedly (see Table 1 for descriptive data), and not all participants endorsed PTSD symptoms. Of the participants with the highest symptom severity, 14 women evidenced fully diagnosed PTSD, and 5 women had significant subthreshold PTSD symptoms that did not meet full criteria (≥ 4 symptoms with ≥ 1 re-experiencing).

Table 1.

Descriptive data for primary study variables

Variables	Mean Score	SD
PTSD Total CAPS Score	25.9	30.1
PTSD CAPS Re-experiencing Score	7.3	8.7
PTSD CAPS Avoidance Score	2.5	4.4
PTSD CAPS Negative Cognitions and Mood	7.8	10.4
PTSD CAPS Arousal Score	8.3	10.0
CO Baseline (l/min)	5.7	1.6
CO Speech Task (l/min)	6.4*	1.6
TPR Baseline (dyne-s/cm5)	1225.8	347.1
TPR Speech Task (dyne-s/cm5)	1246.4	354.9
SBP Baseline (mm Hg)	107.0	11.9
SBP Speech Task (mm Hg)	123.9*	15.0
DBP Baseline (mm Hg)	68.7	8.0
DBP Speech Task (mm Hg)	78.5*	8.9
HR Baseline (bpm)	70.2	11.3
HR Speech Task (bpm)	78.8*	9.9
Dissociation Score	10.5	10.8

^*Significant change from baseline to task (p < .001)

PTSD: Posttraumatic Stress Disorder; CAPS: Clinician Administered Posttraumatic Stress Disorder Scale; CO: Cardiac Output; TPR: total peripheral resistance; SBP: Systolic Blood Pressure; DBP: Diastolic Blood pressure; HR: Heart Rate

One third of the sample reported family income of less than $20,000, another 37% reported in the $20,000-50,000 range, 21% reported 50,000-100,000, and 9% reported over $100,000. With regard to educational attainment, 26% of participants reported high school or equivalent as the terminal degree, 36% reported a Bachelor’s degree, 19% reported Associate’s, 15% reported graduate or professional degrees, and 3% reported “other”.

Primary Analyses

The correlations of CO, TPR, BP and HR reactivity with CAPS derived total PTSD symptom scores and DSM-5 PTSD cluster scores are presented in Table 2. Severity of total PTSD symptom scores was associated with significantly greater CO reactivity (r = .48, p < .01) and lower TPR reactivity (r = −.50, p < .01); these associations did not vary appreciably by PTSD symptom cluster, with correlations for CO reactivity ranging from .40 to .49 across symptom clusters, and correlations for TPR reactivity ranging from −.40 to −.50 across symptom clusters. The PTSD symptom scores were not significantly correlated with SBP reactivity, DBP reactivity, or HR reactivity. No significant correlations were found for the relationship between dissociative symptoms and CVR variables.

Table 2.

Associations of PTSD variables with cardiovascular reactivity measures dnrins sneech task

	CO	TPR	SBP	DBP	HR
PTSD Total	.48*	−.50*	−.23	−.19	−.06
DSM-5 PTSD Cluster B bitrasion symptoms	.42*	−.39*	−.16	−.07	−.10
DSM-5 PTSD Cluster C Avoidance symptoms	.39*	−.38*	−.26	−.17	−.15
DSM-5 PTSD Cluster D Cognitive/Mood symptoms	.48*	−.50	−.15	−.17	.10
DSM-5 PTSD Cluster E Arousal/Reactivity symptoms	.40*	−.48*	−.28	−.27	−.14

^*= p < .01.

CO: cardiac output reactivity; TPR: total peripheral resistance reactivity; SBP: systolic blood pressure reactivity; DBP: diastolic blood pressure reactivity; HR: heart rate reactivity

Discussion

The current study adds to the limited literature on the relationship between PTSD symptom severity and CVR in young civilian women without a history of cardiovascular disease. The use of IC provided a more detailed analysis of the relationships between PTSD symptom clusters and CVR variables during a laboratory stressor. In particular, IC permitted inspection of CO and TPR, the hemodynamic components of BP (Yu et al., 2001; Saab et al., 1992; Van De Water et al., 2003). The present findings did not show any associations of HR, SBP or DBP reactivity with PTSD symptom severity. However, on examination of CO and TPR values, significant associations were found. Our results demonstrated a positive association between PTSD symptom severity and CO reactivity. This finding supports the contention that exaggerated cardiovascular stress responses should be considered as a factor in the increased chronicity in CVD found in women with higher PTSD symptoms (Kubzansky et al., 2009). A negative association was found between PTSD symptom severity and TPR reactivity. Previous research has suggested that this profile of exaggerated cardiac output reactivity, coupled with attenuated vascular resistance, may be observed in relatively healthy and young individuals who are at risk for progressing toward CVD or hypertension (Schneider, Jacobs, Gervirtz, & O’Connor, 2003). This profile has been found to shift toward a greater peripheral resistance response for individuals who develop clinically elevated resting BP. These findings may be salient to the population of relatively healthy young women with greater PTSD symptoms and no current CVD, as we have observed greater BP within the normal range in young women with PTSD (Kibler et al., 2013).

A link between immune system alterations and PTSD may help to explain the progression from attenuated or blunted TPR, to vascular damage in trauma-exposed individuals. (Black & Garbutt, 2002; Tursich et al., 2014). A recent review highlighted a relationship between alterations in inflammatory biomarkers (e.g. pro-inflammatory cytokines, C-reactive protein, fibrinogen) and PTSD (Kibler, Tursich, Ma, Malcolm, & Greenbarg, 2014). Black & Garbutt (2002) suggest that exposure to chronic stress may be associated with inflammatory responses (e.g. increase of cytokines), stress hormone release (e.g. catecholamines, corticosteroids), and lipid oxidation, which may produce damage to the vascular system and increase risk for CVD. Tursich et al. (2014) also found support for this relationship in a recent meta-analysis of 36 studies (n=14,991), reporting moderate correlations between interleukins (IL)-1β, IL-6, tumor necrosis factor (TNF)-α, and C-reactive protein, and trauma exposure. In addition, Dube et al. (2009) provided data suggesting that chronicity of childhood trauma may play a role in immune dysfunction and CVD development. There is a need for further study of this relationship in order to identify early immune dysfunction markers and provide treatment to limit the progression of CVD.

No statistically significant relationships were observed between dissociation and cardiovascular reactivity in the present study. There have been mixed results from previous studies of these relationships. Although studies have associated high dissociation with less reactivity (Halligan et al., 2006; Ladwig et al., 2002; Pole et al., 2005; Sack et al., 2012; Sledjeski & Delahanty, 2012), other studies have found comparable levels of reactivity between high and low dissociators (e.g., Kaufman et al., 2002; Nixon, Bryant, Moulds, Felmingham, & Mastrodomenico, 2005) or no association between dissociation and physiological reactivity (e.g., Schalinski et al., 2014). One factor that may explain our lack of findings relates to the nature of the DES assessment, which was used to measure dissociation in this sample. Several types of dissociation have been described in the literature (e.g., peritraumatic dissociation, persistent dissociation, somatoform dissociation, etc.). The DES quantifies the frequency of dissociative phenomena experienced in daily life (Giesbrecht, Lynn, Lilienfeld, & Merckelbach, 2008), and assesses a trait form of dissociation (Holmes, Oakley, Stuart, & Brewin, 2006). In contrast, other studies that have found a significant relationship between dissociation and reactivity (e.g., Pole et al., 2006; Sledjeski & Delahanty, 2012) have often used measures (e.g., the Peritraumatic Dissociative Experiences Questionnaire; Marmar, Wiess, & Metzler, 1997) that specifically assess for peritraumatic dissociation, a state form of dissociation that occurs during or after a traumatic event. Second, most of the above mentioned studies have used trauma-specific (or trauma-related) imagery, cues, or recall to stress participants prior to assessing dissociation. While the speech task utilized in the present study is successfully used in studies to elicit a cardiovascular stress response, Reinders et al. (2014) note that the use of personal trauma scripts may enhance the breadth of stress responses.

This study was limited by sample size, which did not allow for comparison of cultural and racial differences. Also, our stress reactivity was limited to the speech task, and other varied stressors that emphasize different elements of physiological responding may provide a more robust approach to estimating stress responses. The cross-sectional nature of our study, and lack of immunological data, did not provide an exploration of the interplay between exaggerated or blunted CVR and immune dysfunction in the long-term development of CVD. Future research should include studies of the interplay of stress and immune functioning, in order to develop early CVD risk detection methods and treatment. Despite our limitations, the current study adds to the literature by examining CVR variables in greater detail through the use of impedance-derived CO and TPR, studying a young, healthy cohort of women in order to examine cardiovascular response patterns at a relatively early stage of PTSD, and exploring the impact of PTSD symptom clusters on CVR.