Healthy high-hostiles: reduced parasympathetic activity and decreased sympathovagal flexibility during negative emotional processing

Personality and Individual Differences 36 (2004) 457–469 www.elsevier.com/locate/paid

Healthy high-hostiles: reduced parasympathetic activity and decreased sympathovagal flexibility during negative emotional processing§ Heath A. Demareea,*, D. Erik Everhartb a

Department of Psychology, Case Western Reserve University, Mather Memorial Building, Room 109, 11220 Bellflower Road, Cleveland, OH 44106-7123, USA b Department of Psychology, East Carolina University, Rawl Building, Greenville, NC 27858, USA Received 25 July 2002; received in revised form 3 January 2003; accepted 7 February 2003

Abstract This study was designed to assess the physiological impact of processing neutral, positive, and negative emotional stimuli among a group of low- and high-hostile individuals. Advances were made by (1) measuring both reactivity and recovery to mood induction and (2) using Heart Rate Variability (HRV) to more specifically quantify sympathovagal (Low Frequency divided by High Frequency power, or LF/HF, within the HRV power spectrum) and parasympathetic arousal (Respiratory Sinus Arrhythmia; RSA) at the myocardium. In the present study, men and women were equally divided into low- (N=30) and high-hostile (N=30) groups based on their scores on the Cook–Medley Hostility Scale. Electrocardiogram data were collected before, during, and after being given the negative, positive, or neutral version of the Affective Auditory Verbal Learning Test (AAVL). Results indicate that high-hostiles had reduced parasympathetic activity relative to low-hostiles, as measured by RSA. Moreover, relative to low-hostiles, highhostile participants evidenced reduced sympathovagal reactivity and recovery to the negative AAVL. Results are discussed in terms of their potential value in understanding risk factors for coronary heart disease. # 2003 Elsevier Ltd. All rights reserved. Keywords: Hostility; Emotion; Heart rate variability; Respiratory sinus arrhythmia; Cardiovascular disease; Type A; LF/HF; Heart rate

§

Portions of this manuscript were presented on 5 October 2002, at the 42nd Annual Meeting of the Society for Psychophysiological Research, Washington DC * Corresponding author. Fax: +1-216-368-4891. E-mail addresses: [email protected] (H.A. Demaree), [email protected] (D.E. Everhart). 0191-8869/03/$ - see front matter # 2003 Elsevier Ltd. All rights reserved. doi:10.1016/S0191-8869(03)00109-0

458

H.A. Demaree, D.E. Everhart / Personality and Individual Differences 36 (2004) 457–469

1. Introduction Cardiovascular disease accounts for about 40% of the mortality in the United States as well as in most other industrialized nations (American Heart Association, 2001). The physical, financial, and emotional burdens of this disease on patients, their families and friends are enormous. In fact, the economic cost in the United States alone—which includes payment for medications and medical services, as well as lost income due to death/disability—has been estimated at over $298 billion per year (American Heart Association, 2001). Improving our understanding of the disease’s etiology, including how those predisposed to cardiovascular disease differentially react to stress and emotion, may assist in the prevention and treatment of this malady (American Heart Association, 2001). An abundance of evidence suggests that hostility is positively associated with the risk for coronary heart disease (CHD) and coronary artery disease (CAD) (Brummett & Williams Jr., 1998; Dembroski, MacDougall, Costa, & Grandits, 1989; Littman, 1993; McDermott, Ramsay, & Bray, 2001; Miller, Smith, Turner, Guijarro, & Hallet, 1996; Siegman, Townsend, Civelek, & Blumenthal, 2000). Most recently, McDermott and colleagues (2001) evaluated 97 men with CAD, 28 men with valvular heart disorder without CAD, and 28 men who experienced a bone fracture and were without any heart impairment. Using the latter two groups as controls, these researchers determined that anger expression was a significant predictor of CAD. Dembroski et al. (1989), analyzing data from the prospective Multiple Risk Factor Intervention Trial, found that both Potential for Hostility and an antagonistic interaction style significantly predicted CHD prevalence. Meta-analyses, too, have supported the literature implicating hostility’s role in heart disease. For example, Miller et al. (1996) found that hostility was an independent risk factor for CHD and that the Cook–Medley Hostility Scale (CMHO; Cook & Medley, 1954) was an excellent predictor of ‘‘all-cause mortality’’ (weighted r=0.12) and CHD mortality (weighted r=0.08). One leading theory suggests that increased physiological reactivity to stress leads to increased CHD prevalence among hostile individuals (Williams, Barefoot, & Shekekke, 1985). Although not without debate, this theory has received much support. Numerous studies have found that high-hostiles experience increased physiological reactivity to both everyday and laboratory stressors (Demaree & Harrison, 1997; Demaree, Harrison, & Rhodes, 2000; Edguer, 1994; Everson, McKey, & Lovallo, 1995; Fredrickson, Maynard, Helms, Haney, Siegler, & Barefoot, 2000; Guyll & Contrada, 1998; Lundberg, Hedman, Melin, & Frankenhaeuser, 1989; Powch & Houston, 1996; Rhodes, Harrison, & Demaree, 2002; Smith & Gallo, 1999; Suarez & Williams, 1990; Vogele, 1998). For example, Demaree and Harrison (1997) found that high-hostile men experienced significantly greater heart rate (HR) and systolic blood pressure (SBP) increases to the cold-pressor test relative to their low-hostile counterparts. In research that is perhaps more generalizable, Guyll and Contrada (1998) used ambulatory monitoring devices and found that highhostile men evidenced significantly greater increases of HR and diastolic blood pressure (DBP) during social interaction in comparison to low-hostile men. The literature relating hostility to physiological reactivity is not without contradiction, however. For example, some researchers have failed to find any effect of hostility on cardiovascular parameters to stress (Felsten, 1995; Myrtek, 1995; Suls & Wan, 1993). There are at least two explanations for the contradictory results found in this literature. First, much research in this area has failed to make a distinction between physiological reactivity to a stressor and recovery from that stressor. For instance, Demaree et al. (2000) assessed autonomic functioning while

H.A. Demaree, D.E. Everhart / Personality and Individual Differences 36 (2004) 457–469

459

using ‘‘end-organ’’ physiological measures (HR, SBP, and DBP) both before and after a stressor (the cold-pressor). Although the authors were appropriately cautious when interpreting their results, increased cardiovascular arousal among high-hostiles was generally referred to as ‘‘increased reactivity’’. An equally plausible hypothesis, however, is that low- and high-hostiles experienced similar reactivity but that the latter group evidenced decreased recovery to the stressor. Researchers who have done a better job parsing recovery from reactivity have suggested that a primary physiological consequence of hostility may be decreased recovery (e.g. Ernst, Francis, & Enwonwu, 1990). Second, the vast majority of research has employed the use of ‘‘end-organ’’ cardiovascular measures, such as HR, SBP, and DBP. These measures of the autonomic nervous system are, of course, affected by arousal within both the sympathetic (‘‘speeding’’) and parasympathetic (‘‘slowing’’) branches. The use of end-organ measures increases the probability that altered autonomic arousal goes undetected. For example, comparable arousal increases in both sympathetic and parasympathetic branches may result in no detectable HR change. This investigative problem has received increased attention over the years and several researchers have voiced the need for increased precision in measuring sympathetic and parasympathetic arousal (Beauchaine, 2001; Cacioppo, Uchino, & Berntson, 1994; Gianaros & Quigley, 2001; Rhodes et al., 2002). The use of more discrete autonomic indicators also helps researchers avoid the pitfall of assuming that increased end-organ functioning, such as HR, results from increased sympathetic arousal when, in fact, decreased parasympathetic arousal (i.e., vagal withdrawal) may play at least a partial role. Heart Rate Variability (HRV), used in the present research, is a very important physiological tool because it allows for the relatively discrete measurement of parasympathetic arousal and sympathovagal balance (a significant advantage over ‘‘end-organ measures’’). To perform HRV analyses, an electrocardiograph (ECG) signal is recorded usually using Ag-AgCl electrodes placed on the chest or wrist (for a sample ECG recording, please see Fig. 1). The most noticeable spike in the ECG recording is known as the QRS complex (as denoted in Fig. 1), which occurs once for each heart beat. The first step in HRV analysis is determining the number of milliseconds (ms) between each R-wave, which is known as the interbeat interval (IBI or Heart Period, HP) and is the inverse of HR. The series of IBIs over the duration of recording (5 min in the present study) is then submitted to Fast Fourier Transformation (FFT) to calculate RR-interval spectral power (measured in ms2). In essence, this calculation determines the amount

Fig. 1. Sample ECG, using Mindware Technologies (Westerville, OH) HRV 1.62 computer algorithm.

460

H.A. Demaree, D.E. Everhart / Personality and Individual Differences 36 (2004) 457–469

of IBI change over varied time durations (i.e. Frequency, expressed as Hz). Please see Fig. 2 for a sample FFT output. The area under the FFT curve between 0.15 and 0.40 Hz is known as the High Frequency (HF) component (expressed as ms2/s), and has been found to reflect parasympathetic arousal at the sinoatrial node (SA) of the heart (Berntson, Cacioppo, Binkley, Uchino, Quigley, & Fieldstone, 1994; Eckberg, 1983; Porges, Doussard-Roosevelt, Portales, & Suess, 1994). The reason for this is that changes in IBI corresponding to the human respiratory cycle (0.15–0.40 Hz) are largely controlled by parasympathetic arousal (hence, although highly debated, the term used to express this number is Respiratory Sinus Arrhythmia or RSA). The area under the curve corresponding to the Low Frequency (LF; 0.04–0.15Hz) domain is also measured and expressed in ms2/s. The ratio between absolute LF and absolute HF (LF/HF) is expressed in dimensionless units (Eckberg, 1997). Although a matter of some debate (Berntson et al., 1997; Eckberg, 1997), LF/HF may be thought of as an indicator of sympathovagal balance (Malliani, Pagani, Lombardi, & Cerutti, 1991; Pagani et al., 1986; Task Force, 1996). Despite HRV technology, the impact of hostility on discrete measurements of sympathetic and parasympathetic arousal has received only modest attention. Two studies by Sloan and colleagues (Sloan, Shapiro, Bigger, Bagiella, Steinman, & Gorman, 1994; Sloan et al., 2001), however, found that persons with high levels of hostility (measured by the CMHO) had reduced RSA during waking hours, suggesting that high-hostiles had less vagal modulation at the SA. Because group differences were not significant during sleep, the authors concluded that high hostiles may have experienced more stressful interpersonal interactions during the day (causing decreased parasympathetic arousal) or that they experienced increased vagal withdrawal to psychological stressors (Sloan et al., 2001). Providing support for the latter, these authors found that relative to lowhostiles, high-hostiles experienced enhanced vagal withdrawal to a mental challenge (Sloan et al., 2001). To our knowledge, the remaining research designed to assess the effect of hostility on ANS activity has used pre-ejection period (PEP, a measure of sympathetic arousal, Berntson, Cacioppo, Binkley et al., 1994; Newlin & Levenson, 1979; Sherwood et al., 1997). These studies

Fig. 2. Sample power spectrum, using Mindware Technologies (Westerville, OH) HRV 1.62 computer algorithm.

H.A. Demaree, D.E. Everhart / Personality and Individual Differences 36 (2004) 457–469

461

have largely supported the notion that, relative to their low hostile counterparts, high hostiles respond to stressors with increased sympathetic activity at the SA (Burns, Friedman, & Katkin, 1992; Everson et al., 1995; Sundin & Ohman, 1992). An additional conundrum in this literature is encountered when the role of participant ‘‘harassment’’ is introduced. Specifically, several investigators have determined that negative affect induction (created by giving negative feedback while the participant performs a difficult or stressful task) differentially affects physiological functioning among low- and high-hostiles (Powch & Houston, 1996; Suarez, Harlan, Peoples, & Williams, 1993; Suarez & Williams, 1989). For example, Suarez and Williams (1989) found that experiencing harassment while solving an anagram increased cardiovascular reactivity among high-hostile individuals (as measured by the CMHO). Thus, some tasks that are perhaps more likely to induce negative affect may be more likely to alter cardiovascular functioning among high-hostiles relative to those with less of an emotional component. Indeed, negative and positive emotions have been found to have different physiological consequences. Snyder, Harrison, and Shenal (1998), for example, used the Affective Auditory Verbal Learning test (AAVL) to present their subjects with either negative or positive words to memorize. These researchers found increased and decreased DBP following negative and positive list presentation, respectively. Other researchers have asked participants to verbalize emotional content for mood alteration purposes. Like Snyder and colleagues (1998), several investigators found significantly increased ANS arousal (HR and BP) when saying negatively valenced, as opposed to saying neutral or positively valenced, words (Izard, 1977; McNaughton, 1989; Schwartz, Weinberger, & Singer, 1981; Weerts & Roberts, 1976). Such results suggest that affect, too, plays an important role in physiological arousal. To better assess the impact of hostility and affective valence on physiological reactivity and recovery, the present research was designed to remedy the above problems. First, in addition to an ‘‘end-organ’’ measure [i.e. heart period (HP); the number of milliseconds between each R-wave], procedures designed to better assess sympathovagal balance (the ratio of sympathetic to parasympathetic influences at the heart) and parasympathetic arousal were used. Specifically, ECG data were collected and subsequently analyzed via HRV spectral analysis, yielding both RSA (an excellent indicator of parasympathetic arousal) and LF/HF (an indicator of sympathovagal balance). Second, in order to parse physiological reactivity from recovery, cardiovascular assessment occurred before, during, and after an affective task (as opposed to just before and after). Last, in order to determine the impact of affect, participants experienced a cognitive stressor that had either a neutral, positive, or negative emotional component (the AAVL). 1.1. Hypotheses

Hypothesis 1. Relative to low-hostiles, high hostiles will experience reduced RSA at baseline. Hypothesis 2. High hostiles will evidence increased vagal withdrawal (decreased RSA) and increased sympathovagal balance (LF/HF) to the negative AAVL relative to their low-hostile counterparts. Hypothesis 3. Relative to low-hostiles, high hostiles will experience reduced vagal recovery (smaller increases of RSA) and LF/HF recovery (smaller decreases of LF/HF) to the negative AAVL.

462

H.A. Demaree, D.E. Everhart / Personality and Individual Differences 36 (2004) 457–469

2. Methods 2.1. Variables The classifying variable was self-reported hostility level (low or high) and subjects were also grouped by the valence of the AAVL word list to which they were exposed (negative, positive, and neutral). The dependent measures used in this experiment were HP, RSA, and LF/HF. 2.2. Subjects Subjects included 65 undergraduate men and women from the Introductory Psychology Pool at Case Western Reserve University. Subjects whose data were used in statistical analyses selfreported no history of significant medical problems including seizure, stroke, head trauma, or heart disease. In addition, they reported not currently taking either illicit drugs or prescription medications that may affect their physiological functioning. All subjects received course credit for their participation in this research. 2.3. Self-report Participants were first required to read and sign an Institutional Review Board-approved informed consent form. A demographic/medical questionnaire was given to each subject. While all subjects participated fully in the experiment, data from five subjects were excluded from analyses for medical or drug-related reasons (one for a history of neurological damage and loss of consciousness; two for hypertension and associated medications; and two with hyperthyroidism and associated medications). Data from 60 participants (30 men and 30 women) were included in analyses. Participants were also administered the CMHO (Cook & Medley, 1954), a 50-item true/ false questionnaire that shows validity as a predictor of medical, psychological, and interpersonal outcomes (Han, Weed, Calhoun, & Butcher, 1995). The internal-consistency reliability of the CMHO has been reported as 0.86 (Cook & Medley, 1954). The CMHO cut-off for hostility level for both men and women was 17.5 (‘‘low-hostile’’ <17.5; ‘‘high-hostile’’ > 17.5), which was the median split for our sample. To ensure equal cell sizes, participants in the low- and high-hostile conditions were quasi-randomly placed into the positive, negative, and neutral AAVL conditions (all three affective conditions included 10 men and 10 women, with each sex having a 50% split with regard to hostility level). The experimenter was blind to participant hostility level. 2.4. Apparati 2.4.1. Physiological Participants were tested individually in a quiet (45.000.32 dB) and comfortably lit (about 1300 lux) room. Each ECG data collection period (pre-, during-, and post-AAVL) was 5 min in length, in accordance with the European and American Guidelines (Berntson et al., 1997; Task Force, 1996). Tachogram data were collected at the left and right wrists via Ag/AgCl electrodes and digitized at 1024 samples per second onto a Dell Optiplex GX200 computer. The HP power

H.A. Demaree, D.E. Everhart / Personality and Individual Differences 36 (2004) 457–469

463

spectrum was computed by FFT and the data were linearly de-trended using a computer application created by Biocom Technologies (Biocom, 2000). The average normal NN interval was computed (HP, the inverse of HR with better statistical properties, Berntson, Cacioppo, Quigley, & Fabro, 1994) as well as RSA power (0.15–0.40 Hz, expressed in ms2/sec) and the ratio of LF power (0.04–0.15 Hz) to HF power (i.e., LF/HF). RSA and LF/HF have been found to be good estimates of parasympathetic arousal and sympathovagal balance, respectively. 2.4.2. Perceptual Similar to the format of the Rey Auditory Verbal Learning Test (RAVL) (Rey, 1964), the AAVL consists of two 15-word lists that are to be memorized. However, unlike the RAVL, each of the two AAVL lists has an affective tone (one is positive and one is negative) (Snyder & Harrison, 1997). The positive and negative word lists were developed using an index of word norms (Toglia & Battig, 1978) in which ‘‘pleasantness’’ and ‘‘familiarity’’ were rated on a seven-point Likert scale. The most affective words (positive or negative, as determined by pleasantness ratings) were included in the lists if they rated at least a 5.0/7.0 with regard to familiarity. Moreover, AAVL words were matched to RAVL words with regard to their number of syllables. The RAVL served as the neutral word list. The positive and negative versions of the AAVL have been found to produce reliable, and different, autonomic changes. Moreover, it has also been established that the negative AAVL significantly increases negative affect (Everhart & Demaree, in press). These results suggest that the AAVL may be effective as a mood induction procedure (Snyder et al., 1998) and has been recommended for use in research on emotion (Borod, Tabert, Santschi, & Strauss, 2000). Positive, negative, and neutral word lists were recorded onto TDK IECI/TYPE I audio tapes at a rate of approximately 2 s per word and played at about 50 dB on a Sony CFD-G30 CD Radio Cassette Player. Participant responses were recorded on the same brand of audiocassette using a Radio Shack CTR-122 Cassette Player/Recorder. Subject responses were recorded to aid in data coding. 2.5. Procedure Participants were first seated at a desk and asked to read and sign the informed consent form, as well as to respond to the CMHO and demographic/medical questionnaires. Participants were then seated in a cushioned, upright chair and Biocom Technologies Ag/AgCl electrodes were attached to both wrists (Biocom, 2000). In order to approximate resting physiological conditions, participants were given 10 min to adjust to the laboratory environment. Prior to ECG data collection, the researcher gave the following instructions: ‘‘During this experiment, the electrodes on your wrists will collect physiological data from your body. The electrodes work best when they do not slip on your skin, which is often caused by movement. Accordingly, please get into a comfortable, relaxed position and try to maintain that position during the experiment.’’ The researcher then initiated the first 5-min ECG data collection. The next 5-min ECG data collection period occurred during AAVL stimulus presentation for mood induction purposes (either the positive, negative, or neutral version was presented). The AAVL word list was presented across five trials. Prior to the first trial, the participant received the following audiotaped instructions: ‘‘I am going to read you a list of words. Please listen carefully. When I stop, please say as many words as you can remember. Say the words in any order. Just say as many as you can.’’ For Trials 2–5, the participant received these instructions: ‘‘Now I am going to read the

464

H.A. Demaree, D.E. Everhart / Personality and Individual Differences 36 (2004) 457–469

same list again. When I stop, I want you to tell me as many words as you can remember including the words you have said before. It doesn’t matter what order you say them in. Just say as many words as you can.’’ Subsequent to each Trial’s instructions, the participant heard the list of words (played over audiotape) and then responded with the words that they could remember. To assess ‘‘recovery,’’ a final 5-min ECG data collection period followed the AAVL administration. It should be noted that the 5-trial AAVL administration often took longer than 5 min. Regardless, recovery ECG data collection ensued immediately after they indicated that they could not recall any more words. Subjects were then debriefed and thanked for their participation.

3. Results To compare groups (low- vs. high-hostility) on descriptive measures, t-tests were conducted on age and CMHO. As expected, low- and high-hostile groups differed significantly with regard to CMHO score (Low-Hostile: M=13.40, S.D.=3.90; High-Hostile: M=23.33, S.D.=4.16) [t(58)=9.55, P<0.01] but not age (Low-Hostile: M=18.43 years, S.D.=0.77; High-Hostile: M=19.03 years, S.D.=2.31) [t(58)=1.35, P>0.05]. Although recent research suggests otherwise (e.g. Denver & Porges, 2001), verbal output (affecting respiration) may affect HRV indices (Berntson et al., 1997; Task Force, 1996). Thus, it is important to ensure that no group differences existed with regard to the number of AAVL words recalled. No Hostility group (Low-Hostile: M=51.63 words, S.D.=7.51; High-Hostile: M=52.53 words, S.D.=7.34) [t(58)=0.47, P>0.05] or Sex group differences (Men: M=51.57 words, S.D.=7.61; Women: M=52.59 words, S.D.=6.59) [t(58)=0.64, P>0.05] were found with regard to total number of words recalled. 3.1. Hypotheses Independent analyses of variance (ANOVAs) were performed on three dependent physiological variables—HP, RSA, and LF/HF. Each ANOVA was performed with independent factors of Hostility and Valence and a repeated measure of Phase (Pre-, During-, and Post-AAVL). ‘‘Reactivity’’ was quantified by subtracting baseline data from corresponding values collected during AAVL administration. Similarly, ‘‘recovery’’ was quantified by subtracting data collected during AAVL administration from values collected following AAVL administration. All pairwise comparisons were made using Tukey’s Least Significant Difference test to control for experimentwise error rate (Winer, 1971). For HP, analysis revealed only a significant main effect of Phase [F(2, 54)=55.33, P< 0.01]. Specifically, HP was shorter during the AAVL (M=776.36 ms; S.D.=121.90) relative to before (M=835.77 ms; S.D.=128.08) or after (M=839.85 ms; S.D.=119.92) the AAVL. Analysis of RSA data, an indicator of parasympathetic arousal, revealed a significant main effect of Hostility [F(1, 54)=5.17, P<0.05]. Specifically, collapsed across all three phases, highhostiles had reduced RSA levels (M=319.70 ms2/s; S.D.=345.59) relative to low-hostiles (M=515.60 ms2/s; S.D.=378.29). Post-hoc analyses also revealed significant group differences at baseline (Low-Hostile: M=554.52 ms2/s, S.D.=387.10; High-Hostile: M=352.26 ms2/s, S.D.=422.46), even before any AAVL presentation. It is important to note that 76.7% (23/30) of

465

H.A. Demaree, D.E. Everhart / Personality and Individual Differences 36 (2004) 457–469

high-hostile participants evidenced RSA levels lower than the low-hostile RSA mean (i.e. 554.52 ms2/ s) at baseline and that no RSA data point was deemed an outlier (as defined as the mean 3 S.D.s). In addition, a significant main effect of Phase was also found [F(2, 54)=4.13, P<0.05] in which RSA during AAVL presentation (M=360.10 ms2/s; S.D.=279.20) was significantly lower than either before (M=453.39 ms2/s; S.D.=414.46) or after (M=439.46 ms2/s; S.D.=412.81) the AAVL. For LF/HF data (sympathovagal balance), a significant interaction of Hostility by Valence by Phase was found [F(6, 106)=4.13, P<0.05]. LF/HF data are displayed in Table 1. Post-hoc analyses revealed that high-hostiles (M=3.40; S.D.=2.66) evidenced significantly higher LF/HF levels at baseline relative to their low-hostile counterparts (M=2.36; S.D.=1.81). Further, low-hostiles experienced significant LF/HF increases during both reactivity and recovery phases to the negative AAVL whereas high-hostiles showed no significant changes in LF/HF. To neutral AAVL exposure, the low-hostile group evidenced significantly increased LF/HF during recovery whereas the high-hostile group showed no significant changes in LF/HF during either reactivity or recovery phases.

4. Discussion The present experiment was designed, in part, to replicate prior research which has found decreased parasympathetic arousal and increased physiological reactivity to stress among highhostile individuals. Moreover, in an attempt to extend previous work, the present research design included (1) discrete measures of parasympathetic arousal and sympathovagal balance (as opposed to end organ measures), (2) psychological stressors with three affective components (positive, negative, and neutral), and (3) measurement of both reactivity and recovery. The primary and most robust finding in the present experiment is that high-hostile individuals have reduced parasympathetic arousal relative to low-hostiles at baseline, thus supporting Hypothesis 1. To our knowledge, this is the first replication of this finding outside of the Sloan laboratory (Sloan et al., 1994, 2001). Interestingly, post-hoc analyses revealed a significant difference between low- and high-hostiles prior to AAVL presentation. Given that all subjects were given 10 min to become accustomed to the laboratory environment, these group differences may reflect differential baseline autonomic functioning rather than a confound of increased high-hostile reactivity to mere laboratory exposure. The results warrant the consideration that CHD may additionally be driven by elevated cardiovascular arousal at rest. This result, in juxtaposition to Table 1 LF/HF ratio data by hostility, valence, and phase

Low-Hostile Positive High-Hostile Positive Low-Hostile Negative High-Hostile Negative Low-Hostile Neutral High-Hostile Neutral

Pre-

During-

Post-

2.34 (1.85) 3.80 (3.17) 1.92 (1.27) 3.19 (1.93) 2.82 (2.35) 3.22 (2.89)

2.99 (1.84) 3.81 (2.35) 2.68 (1.73) 3.62 (1.73) 2.13 (1.07) 2.79 (1.16)

2.99 5.11 3.28 3.59 2.90 3.09

(1.65) (3.54) (2.84) (2.16) (2.21) (1.27)

466

H.A. Demaree, D.E. Everhart / Personality and Individual Differences 36 (2004) 457–469

nonsignificant low- and high-hostile group differences on the end-organ measurement of HP, highlights the importance of using relatively discrete measurements of sympathovagal balance and parasympathetic arousal to assess for individual differences in autonomic functioning (as well as the prediction of heart disease; Beauchaine, 2001; Cacioppo et al., 1994; Gianaros & Quigley, 2001; Rhodes et al., 2002). Hypotheses 2 and 3, which predicted increased reactivity and decreased recovery of RSA and LF/ HF measures among high-hostiles to the negative AAVL, were not supported. Specifically, both lowand high-hostile groups evidenced reduced RSA to the negative AAVL and only the low-hostile group showed increased LF/HF to the negative AAVL. For recovery, both hostility groups evidenced increased RSA following the negative AAVL but the low-hostility group simultaneously showed increased LF/HF! The finding of increased LF/HF among low-hostiles during negative AAVL recovery was surprising, and group differences might be explained by the Law of Initial Values (Furedy & Scher, 1989). This Law posits that, because of tendencies for data to regress towards the overall mean, relatively low values (in this case, LF/HF values of low-hostiles) have an increased probability of increasing rather than decreasing. That is, increased LF/HF among low-hostiles during recovery may have resulted from their having reduced sympathovagal balance premorbidly. It is also interesting to note that, compared to low-hostiles, high-hostiles responded with significantly reduced LF/HF reactivity and recovery changes to both the neutral and negative AAVL. Decreased sympathovagal balance—associated with ‘‘parasympathetic flexibility’’ or ‘‘spectral reserve’’—has been consistently found in persons with anxiety or panic (Friedman & Thayer, 1998a, 1998b; Thayer, Friedman, & Borkovec, 1996; Thayer, Smith, Rossy, Sollers, & Friedman, 1998). Interestingly, anxiety has also been implicated as a risk factor for CHD (Fava, Abraham, Pava, Shuster, & Rosenbaum, 1996; Winters & Schneiderman, 2000). Taken together, these results suggest that reduced RSA and increased LF/HF at rest, as well as decreased ANS ‘‘flexibility’’ to emotional stressors (perhaps due to a parasympathetic ‘‘floor’’ effect), may be risk factors for CHD. The authors identify three main weaknesses in the present study. First, while LF power is most frequently associated with sympathetic arousal, this relationship has been difficult to establish and thus LF and LF/HF should be interpreted with caution (Berntson et al., 1997; Task Force, 1996). Future research should consider using RSA as an indicator of parasympathetic arousal (which has been well established) along with a better indicator of sympathetic activity, such as PEP (e.g. Beauchaine, 2001; Berntson, Cacioppo, Binkley et al., 1994; Newlin & Levenson, 1979; Sherwood et al., 1997). Second, phasic differences may partially result from altered respiration due to vocal output. Thus, researchers may wish to consider the use of emotional stimuli that induce affective responses with minimal peripheral changes (e.g. films or slides). Last, although the AAVL has been found to produce reliable effects on self-reported mood (Everhart & Demaree, in press), future research should incorporate a manipulation check for mood induction. References American Heart Association. (2001). Heart and stroke and statistical update. Dallas Texas: American Heart Association. Beauchaine, T. (2001). Vagal tone, development, and Gray’s motivational theory: toward an integrated model of autonomic nervous system functioning in psychopathology. Development and Psychopathology, 13(2), 183–214. Berntson, G. G., Bigger, J. T. Jr., Eckberg, D. L., Grossman, P., Kaufmann, P. G., Malik, M., Nagaraja, H. N.,

H.A. Demaree, D.E. Everhart / Personality and Individual Differences 36 (2004) 457–469

467

Porges, S. W., Saul, J. P., Stone, P. H., & van der Molen, M. W. (1997). Heart rate variability: origins, methods, and interpretive caveats. Psychophysiology, 34(6), 623–648. Berntson, G. G., Cacioppo, J. T., Binkley, P. F., Uchino, B. N., Quigley, K. S., & Fieldstone, A. (1994). Autonomic cardiac control. III. Psychological stress and cardiac response in autonomic space as revealed by pharmacological blockades. Psychophysiology, 31(6), 599–608. Berntson, G. G., Cacioppo, J. T., Quigley, K. S., & Fabro, V. T. (1994). Autonomic space and psychophysiological response. Psychophysiology, 31(1), 44–61. Biocom. (2000). Heart rhythm scanner. Washington: Biocom Technologies. Borod, J. C., Tabert, M. H., Santschi, C., & Strauss, E. H. (2000). Neuropsychological assessment of emotional processing in brain-damaged patients. In J. C. Borod (Ed.), The neuropsychology of emotion (pp. 80–105). New York: Oxford University Press. Brummett, B. H., & Williams, R. B. Jr (1998). Hostility and risk for disease. Current Opinion in Psychiatry, 11, 607– 613. Burns, J. W., Friedman, R., & Katkin, E. S. (1992). Anger expression, hostility, anxiety, and patterns of cardiac reactivity to stress. Behavioral Medicine, 18(2), 71–78. Cacioppo, J. T., Uchino, B. N., & Berntson, G. G. (1994). Individual differences in the autonomic origins of heart rate reactivity: the psychometrics of respiratory sinus arrhythmia and preejection period. Psychophysiology, 31(4), 412–419. Cook, W. W., & Medley, D. M. (1954). Proposed hostility and paraphaisic-virtue scales for the MMPI. The Journal of Applied Psychology, 38(6), 414–418. Demaree, H. A., & Harrison, D. W. (1997). Physiological and neuropsychological correlates of hostility. Neuropsychologia, 35(10), 1405–1411. Demaree, H. A., Harrison, D. W., & Rhodes, R. D. (2000). Quantitative electroencephalographic analyses of cardiovascular regulation in low- and high-hostile men. Psychobiology, 28(3), 420–431. Dembroski, T. M., MacDougall, J. M., Costa, P. T. Jr, & Grandits, G. A. (1989). Components of hostility as predictors of sudden death and myocardial infarction in the Multiple Risk Factor Intervention Trial. Psychosomatic Medicine, 51(5), 514–522. Denver, J. W., & Porges, S. W. (2001). Methodologic issues in the quantification of RSA [abstract]. Psychophysiology, 38(Supplement 1), S35. Eckberg, D. L. (1983). Human sinus arrhythmia as an index of vagal cardiac outflow. Journal of Applied Physiology: Respiratory and Environmental Exercise Physiology, 54, 961–966. Eckberg, D. L. (1997). Sympathovagal balance: a critical appraisal. Circulation, 96(9), 3224–3232. Edguer, N. (1994). Type A behaviour and aggression: provocation, conflict and cardiovascular responsivity in the Buss teacher-learner paradigm. Personality and Individual Differences, 17(3), 377–393. Ernst, F. A., Francis, R. A., & Enwonwu, C. O. (1990). Manifest hostility may affect habituation of cardiovascular reactivity in blacks. Behavioral Medicine, 16(3), 119–124. Everhart, D. E., & Demaree, H. A. Low alpha power (7.5–9.5 Hz) changes during positive and negative affective learning. Cognitive, Affective, and Behavioral Neuroscience (in press). Everson, S. A., McKey, B. S., & Lovallo, W. R. (1995). Effect of trait hostility on cardiovascular responses to harrassment in young men. International Journal of Behavioral Medicine, 2(2), 172–191. Fava, M., Abraham, M., Pava, J., Shuster, J., & Rosenbaum, J. (1996). Cardiovascular risk factors in depression: the role of anxiety and anger. Psychosomatics: Journal of Consultation Liasion Psychiatry, 37(1), 31–37. Felsten, G. (1995). Cynical hostility influences anger, but not cardiovascular reactivity during competition with harassment. International Journal of Psychophysiology, 19(3), 223–231. Fredrickson, B. L., Maynard, K. E., Helms, M. J., Haney, T. L., Siegler, I. C., & Barefoot, J. C. (2000). Hostility predicts magnitude and duration of blood pressure response to anger. Journal of Behavioral Medicine, 23(3), 229–243. Friedman, B. H., & Thayer, J. F. (1998a). Anxiety and autonomic flexibility: a cardiovascular approach. Biological Psychology, 47(3), 243–263. Friedman, B. H., & Thayer, J. F. (1998b). Autonomic balance revisited: panic anxiety and heart rate variability. Journal of Psychosomatic Research, 44(1), 133–151. Furedy, J. J., & Scher, H. (1989). The law of initial values: differentiated testing as an empirical generalization versus enshrinement as a methodological rule. Psychophysiology, 26, 120–121.

468

H.A. Demaree, D.E. Everhart / Personality and Individual Differences 36 (2004) 457–469

Gianaros, P. J., & Quigley, K. S. (2001). Autonomic origins of a nonsignal stimulus-elicited bradycardia and its habituation in humans. Psychophysiology, 38(3), 540–547. Guyll, M., & Contrada, R. J. (1998). Trait hostility and ambulatory cardiovascular activity: responses to social interaction. Health Psychology, 17(1), 30–39. Han, K., Weed, N. C., Calhoun, R. F., & Butcher, J. N. (1995). Psychometric characteristics of the MMPI-2 CookMedley Hostility Scale. Journal of Personality Assessment, 65(3), 567–585. Izard, C. E. (1977). Human emotions. New York: Plenum Press. Littman, A. B. (1993). Review of psychosomatic aspects of cardiovascular disease. Psychotherapy and Psychosomatics, 60(3-4), 148–167. Lundberg, U., Hedman, M., Melin, B., & Frankenhaeuser, M. (1989). Type A behavior in healthy males and females as related to physiological reactivity and blood lipids. Psychosomatic Medicine, 51(2), 113–122. Malliani, A., Pagani, M., Lombardi, F., & Cerutti, S. (1991). Cardiovascular neural regulation explored in the frequency domain. Circulation, 84, 482–492. McDermott, M. R., Ramsay, J. M. C., & Bray, C. (2001). Components of the anger-hostility complex as risk factors for coronary artery disease severity: a multi-measure study. Journal of Health Psychology, 6(3), 309–319. McNaughton, N. (1989). Biology and emotion. New York: Cambridge University Press. Miller, T. Q., Smith, T. W., Turner, C. W., Guijarro, M. L., & Hallet, A. J. (1996). A meta-analytic review of research on hostility and physical health. Psychological Bulletin, 119(2), 322–348. Myrtek, M. (1995). Type A behavior pattern, personality factors, disease, and physiological reactivity: a meta-analytic update. Personality and Individual Differences, 18(4), 491–502. Newlin, D. B., & Levenson, R. W. (1979). Pre-ejection period: measuring beta-adrenergic influences upon the heart. Psychophysiology, 16(6), 546–553. Pagani, M., Lombardi, F., Guzzetti, S., Rimoldi, O., Furlan, R., Pizzinelli, P., Sandrone, G., Malfatto, G., Dell’Orto, S., & Piccaluga, E. (1986). Power spectral analysis of heart rate and arterial pressure variabilities as a marker of sympatho-vagal interaction in man and conscious dog. Circulation Research, 59, 178–193. Porges, S. W., Doussard-Roosevelt, J. A., Portales, A. L., & Suess, P. E. (1994). Cardiac vagal tone: Stability and relation to difficultness in infants and 3-year-olds. Developmental Psychobiology, 27(5), 289–300. Powch, I. G., & Houston, B. K. (1996). Hostility, anger-in, and cardiovascular reactivity in white women. Health Psychology, 15(3), 200–208. Rey, A. (1964). L’examin clinique en psychologie. Paris: Universitaire de France. Rhodes, R. D., Harrison, D. W., & Demaree, H. A. (2002). Hostility as a moderator of physiological reactivity and recovery to stress. International Journal of Neuroscience, 112, 167–186. Schwartz, G. E., Weinberger, D. A., & Singer, J. A. (1981). Cardiovascular differentiation of happiness, sadness, anger, and fear following imagery and exercise. Psychosomatic Medicine, 43, 343–364. Sherwood, A., Girdler, S. S., Bragdon, E. E., West, S. G., Brownley, K. A., Hinderliter, A. L., & Light, K. C. (1997). Ten-year stability of cardiovascular responses to laboratory stressors. Psychophysiology, 34(2), 185– 191. Siegman, A. W., Townsend, S. T., Civelek, A. C., & Blumenthal, R. S. (2000). Antagonistic behavior, dominance, hostility, and coronary heart disease. Psychosomatic Medicine, 62(2), 248–257. Sloan, R. P., Bagiella, E., Shapiro, P. A., Kuhl, J. P., Chernikhova, D., Berg, J., & Myers, M. M. (2001). Hostility, gender, and cardiac autonomic control. Psychosomatic Medicine, 63, 434–440. Sloan, R. P., Shapiro, P. A., Bigger, J. T. Jr., Bagiella, E., Steinman, R. C., & Gorman, J. M. (1994). Cardiac autonomic control and hostility in healthy subjects. American Journal of Cardiology, 74(3), 298–300. Smith, T. W., & Gallo, L. C. (1999). Hostility and cardiovascular reactivity during marital interaction. Psychosomatic Medicine, 61(4), 436–445. Snyder, K. A., & Harrison, D. W. (1997). The Affective Auditory Verbal Learning Test. Archives of Clinical Neuropsychology, 12(5), 477–482. Snyder, K. A., Harrison, D. W., & Shenal, B. V. (1998). The Affective Auditory Verbal Learning Test: peripheral arousal correlates. Archives of Clinical Neuropsychology, 13(3), 251–258. Suarez, E. C., Harlan, E., Peoples, M. C., & Williams, R. B. Jr. (1993). Cardiovascular and emotional responses in women: the role of hostility and harassment. Health Psychology, 12(6), 459–468.

H.A. Demaree, D.E. Everhart / Personality and Individual Differences 36 (2004) 457–469

469

Suarez, E. C., & Williams, R. B. Jr. (1989). Situational determinants of cardiovascular and emotional reactivity in high and low hostile men. Psychosomatic Medicine, 51(4), 404–418. Suarez, E. C., & Williams, R. B. Jr. (1990). The relationships between dimensions of hostility and cardiovascular reactivity as a function of task characteristics. Psychosomatic Medicine, 52(5), 558–570. Suls, J., & Wan, C. K. (1993). The relationship between trait hostility and cardiovascular reactivity: a quantitative review and analysis. Psychophysiology, 30(6), 615–626. Sundin, O., & Ohman, A. (1992). Cardiovascular response patterning and type A behavior in middle aged men. Journal of Psychophysiology, 6, 240–251. Task Force. (1996). Heart rate variability. Standards of measurement, physiological interpretation, and clinical use. European Heart Journal, 17(3), 354–381. Thayer, J. F., Friedman, B. H., & Borkovec, T. D. (1996). Autonomic characteristics of generalized anxiety disorder and worry. Biological Psychiatry, 39(4), 255–266. Thayer, J. F., Smith, M., Rossy, L. A., Sollers, J. J., & Friedman, B. H. (1998). Heart period variability and depressive symptoms: gender differences. Biological Psychiatry, 44(4), 304–306. Toglia, M. P., & Battig, W. F. (1978). Handbook of word norms. Hillsdale, NJ: Lawrence Erlbaum. Vogele, C. (1998). Serum lipid concentrations, hostility and cardiovascular reactions to mental stress. International Journal of Psychophysiology, 28(2), 167–179. Weerts, T. C., & Roberts, R. (1976). The psychophysiological effects of imagining anger-provoking and fear-provoking scenes. Psychophysiology, 13, 174. Williams, R. B. Jr, Barefoot, J. C., & Shekekke, R. B. (1985). The health consequences of hostility. In M. A. Chesney, & R. H. Rosenman (Eds.), Anger and hostility in cardiovascular and behavioral disorders? (pp. 173–185). Washington, DC: Hemisphere. Winer, B. J. (1971). Statistical principles in experimental design. New York: McGraw-Hill. Winters, R. W., & Schneiderman, N. (2000). Anxiety and coronary heart disease. In D. I. Mostofsky, & D. H. Barlow (Eds.), The management of stress and anxiety in medical disorders (pp. 206–219). New York: Allyn and Bacon.