خصومت، ابراز هیجانی و واکنش همودینامیک به عوامل استرس زای آزمایشگاهی:اثرات واکنش شدت تمایل به ابراز احساسات برای همگان

Abstract The present study examined the interactive effects of hostility and a predisposition towards emotional expression or suppression in interpersonal situations. We also attempted to partially replicate findings from a recent investigation which provided evidence of lower myocardial and greater vascular responses in high-hostile relative to low-hostile individuals. Undergraduate students (n = 99) participated in a protocol consisting of rest periods, speech preparation and presentation, a social-evaluative mental arithmetic task, and a stress interview. After classifying participants into high/low hostility by high/low interpersonal emotional expression groups using median-splits, high-hostile individuals showed lower HR and SBP responses to speech preparation and reported greater threat appraisal and negative affect than low-hostile participants. High-hostile interpersonal expressors and male interpersonal expressors displayed lower DBP and TPR reactivity, respectively, than high-hostile or male suppressors. High-hostile expressors also reported lower levels of positive affect than high-hostile suppressors, and expressors reported lower threat appraisals than suppressors, irrespective of hostility. Findings are discussed in terms of an absence of conflict or ambivalence over interpersonal emotional expression for high-hostile expressors and are suggestive of potential health benefits of expressing emotion interpersonally for at-risk groups.

Introduction A considerable body of evidence supports an association between cynical hostility [assessed by the Cook-Medley Hostility Scale (Ho; Cook and Medley, 1954)] and an increased risk of coronary heart disease (CHD) (for reviews, see Miller et al., 1996 and Smith, 1992). Cynical hostility, defined as “a set of negative attitudes, beliefs, and appraisals concerning others…[including]…a belief that others are generally unworthy and not to be trusted” (Smith, 1992, p. 139), has been described as an interpersonal risk factor (e.g., Smith and Gerin, 1998). In reviewing the relevant literature, Timothy Smith (1992) found support for the idea that “…[hostile] individuals…experience a more taxing interpersonal environment” (p. 145). According to Smith (1992), the “transactional model” of the hostility–health relationship assumes that hostile people are likely to appraise the behavior of others as indicative of “hostile intent” (p. 145) and “…to elicit and exacerbate conflict in their daily lives” (p. 145). Consistent with this view, recent research has found that individuals with high hostility appraised videotapes depicting ambiguous social interactions more negatively ( Vranceanu et al., 2006). The idea that such interpersonal cynicism and hypersensitivity connotes risk is supported by evidence that stress-induced cardiovascular reactivity, a putative mechanism for the hostility-CHD link, is reliably associated with hostility in the laboratory only when social/interpersonal stressors are examined (for reviews, see Smith, 1992 and Suls and Wan, 1993). Given their difficulty trusting others, the prospect of expressing emotions interpersonally might be especially daunting for some hostile individuals. Consistent with this notion, Christensen and Smith (1993) found an association between hostility and heightened blood pressure reactivity among men who were asked to disclose personal details to a stranger. However, while situational self-disclosure undoubtedly involves some degree of interpersonal emotional expression, participants' dispositional tendencies toward expressing versus suppressing their emotions interpersonally were not examined. This attribute could be particularly important in terms of disease relevance as it has been observed that hypertensive patients are characterized by internal conflict between the experience of hostility or anger and a desire to avoid expressing emotion to others (Alexander, 1939). Indeed, it is becoming increasingly evident that chronic emotional suppression carries risk for illness in its own right, whereas expression promotes well-being (e.g., for reviews, see Pennebaker, 1989, Pennebaker, 1997 and Scheier and Bridges, 1995). Thus, cynical hostility, along with a tendency to suppress emotion in interpersonal situations, in conjunction with life experiences that require the contrary, might be a particularly toxic combination. Conversely, the capacity to express emotion interpersonally might serve to buffer or ameliorate the potentially harmful effects of hostility. Most research examining relationships between cardiovascular reactivity and emotional expression, alone or in conjunction with the experience of hostility, has focused on the expression of specific negative emotions, most notably, anger (for review, see Siegman, 1993). Although findings have been mixed (e.g., Bongard et al., 1998, Burns, 1995 and Vögele et al., 1997), several investigations have found trait anger expression to be related to greater reactivity than suppression, but this relationship consistently emerges only when laboratory conditions are designed to intentionally induce anger (e.g., Siegman, 1993). Furthermore, in addition to emphasizing a single negative emotion, these studies often define anger expression using measures such as Spielberger's (1991) Anger-In and Anger-Out scales, which include items (e.g., “I do things like slam doors”) that do not seem to tap interpersonal expression. Investigations not restricted to the study of anger have tended to examine situation-specific expression of primarily negative emotions ( Mendes et al., 2003), rather than the dispositional tendency to express positive and negative emotions in general. Individuals differ markedly in the extent to which their blood pressure responses to one or more stressors are driven primarily by increases in cardiac output (CO), total peripheral resistance (TPR), or both (Manuck et al., 1993) and these hemodynamic response patterns, known as myocardial, vascular, and mixed responses, respectively, may have differential pathophysiological implications (Obrist, 1981 and Saab and Schneiderman, 1993). While longitudinal data are still needed, there is increasing evidence suggesting that vascular responses (Goldberg et al., 1996, Kapuku et al., 1999, Sherwood et al., 1999 and Treiber et al., 1993) may be more maladaptive than myocardial responses (Heponiemi et al., 2007). Despite the apparent meaningfulness of interindividual variability in hemodynamic profiles, surprisingly few studies have attempted to relate hostility (with or without anger expression/suppression) to CO or TPR reactivity. While Bongard et al. (1998) did not find evidence of such relationships, high-hostile anger expressors tended to show the least overall blood pressure reactivity, while low-hostile anger expressors tended to be most reactive. In a more recent investigation by Davis et al. (2000), high-hostile participants exhibited less heart rate (HR), systolic blood pressure (SBP), and CO reactivity and greater diastolic blood pressure (DBP) and TPR responses than their low-hostile counterparts. Associations of psychosocial traits like hostility with cardiovascular responses may be due, in part, to stressor-related appraisals, emotions, and/or behaviors habitually experienced by people with the trait in question. In terms of appraisals, perhaps the best candidates in this regard are challenge and threat appraisals, which have been shown to precede myocardial and vascular/mixed responses to laboratory stressors, respectively (Tomaka et al., 1993 and Tomaka et al., 1997). While data are remarkably lacking in this regard, it seems reasonable to predict that high-hostile individuals might be predisposed to appraise social/interpersonal stressors as more threatening than their low-hostile counterparts. This would seem to be particularly true of hostile people who are uncomfortable with expressing their emotions to others. For example, although Mendes et al. (2003; Study 2) conceptualized challenge and threat as situational emotional expression and suppression (to same-sex research assistants), respectively, as opposed to defining them in terms of appraisals, they observed cardiovascular sequelae commonly associated with the latter. The present study was intended to partially replicate the findings of Davis et al. (2000) by examining hemodynamic responses to subtle social stressors as a function of hostility. We also sought to extend the existing literature by exploring the interactive effects of hostility and an index of interpersonal expressive style on hemodynamic responsivity, appraisals, and state affect.

Results 3.1. Demographic and baseline characteristics Demographic and baseline characteristics of high/low hostility by high/low interpersonal emotional expression groups are shown in Table 1. Chi-square tests of independence conducted to examine potential hostility by expression group differences in gender composition, χ2 (3, N = 99) = 5.26, p > .10, and smoking status, χ2 (3, N = 97) = 3.36, p > .10, as well as ethnic composition and parental history of hypertension, χ2s (6, N = 99) < 4.23, ps > .10, yielded no group differences. Due to small cell sizes in some of the above analyses and the possibility of hostility or emotional expression group differences irrespective of the other variable, chi-square analyses were also performed separately for high/low hostility groups and high/low emotional expression groups. These analyses yielded a trend toward a difference in the gender composition of hostility groups that was nearly significant, χ2 (1, N = 99) = 3.73, p = .053, with the low-hostile group being composed of a greater percentage of women and a smaller percentage of men than the high-hostile group. Gender was included as a control variable in all remaining analyses. Table 1. Demographic and baseline characteristics of high/low hostility × high/low interpersonal emotional expression groups Variable Low HOST–Low EX Low HOST–High EX High HOST–Low EX High HOST–High EX Gender Men 41% 24% 47% 53% Women 59% 76% 53% 47% Ethnicity African-American 12% 33% 20% 21% Caucasian 70% 55% 60% 68% Other 18% 12% 20% 11% PH Positive 53% 58% 53% 53% Negative 18% 24% 20% 31% Uncertain 29% 18% 27% 16% Smoking status Smokers 24% 15% 14% 33% Nonsmokers 76% 85% 86% 67% Age (yr) 20.1 (4.7) 24.8(8.4) 20.3 (3.5) 21.1 (3.4) Height (in) 66.1 (4.5) 65.4 (4.0) 66.3 (3.4) 67.1 (4.1) Weight (lb) 151.1(50.9) 162.0 (40.7) 158.3 (28.6) 160.9 (43.0) BMI (kg/m2) 23.9 (6.0) 26.5(6.0) 25.2 (3.6) 25.0 (5.5) BSA (m2) 1.76(.31) 1.81(.24) 1.81(.19) 1.83 (.26) CASSBPa (mm Hg) 113 (13.0) 114 (8.7) 114 (7.3) 112 (10.1) CASDBPa (mm Hg) 69 (8.8) 70 (9.0) 71 (9.6) 72 (8.3) SBPb (mm Hg) 108.4(13.4) 108.7(9.1) 110.7 (8.7) 109.7 (11.6) DBPb (mm Hg) 61.6(7.4) 61.0 (7.4) 61.3 (6.1) 62.9 (6.9) HRb (bpm) 74.2(10.0) 72.7(10.5) 74.9(11.7) 73.0(8.7) COb (L/min) 5.4(1.18) 5.7(1.03) 5.9(1.09) 5.9 (1.19) TPRb (pru) .89(.21) .83 (.15) .85(.16) .84 (.17) Values for age, height, weight, BMI, BSA, CASSBP, CASDBP, SBP, DBP, HR, CO, and TPR are unadjusted means (SD). HOST—hostility; EX—interpersonal emotional expression; PH—parental history of hypertension; BMI—body mass index; BSA—body surface area; CASSBP—casual SBP; CASDBP—casual DBP; SBP—systolic blood pressure; DBP—diastolic blood pressure; HR—heart rate; CO—cardiac output; TPR—total peripheral resistance. a Casual blood pressure assessed via mercury sphygmomanometer. b Initial resting levels of physiological variables. Table options A 2 (Hostility Group) × 2 (Interpersonal Expression Group) × 2 (Gender) ANOVA revealed a significant main effect of expression group for age, with interpersonal expressors being older than suppressors, F(7,91) = 6.64, p = .012. Age was included as a covariate in all remaining analyses. A series of 2 (Hostility Group) × 2 (Interpersonal Expression Group) × 2 (Gender) ANCOVAs performed on the remaining variables in Table 1 yielded no significant main effects or interactions involving hostility or emotional expression group, Fs(8, 90) < 1.58, ps > .10. 1 While no group differences were observed for BMI, it should be noted that 30% of the participants were overweight (25 < BMI < 29.9) and 16% were obese (BMI > 30) ( National Heart, Lung, and Blood Institute, 2006). 3.2. Hostility × interpersonal emotional expression group differences in reactivity The mixed ANCOVA for CO revealed no main effects of hostility or emotional expression group or interactions among hostility group, expression group (see Table 2), and/or gender, Fs(1,90) < 3.05, ps > .10. There were also no interactions involving hostility and/or expression group with task, Fs(3,270) < 1.45, ps > .10. Table 2. Changes from pretask baseline to task, averaged across stressors, for high/low hostility × high/low interpersonal emotional expression groups Variable Low HOST–Low EX Low HOST–High EX High HOST–Low EX High HOST–High EX SBP (mm Hg) 13.7 (8.1) 14.6 (8.0) 14.3 (6.8) 10.3 (6.5) DBP (mm Hg) 9.0 (5.2) 8.9 (5.1) 9.5 (3.6) 5.3(3.9) HR (bpm) 10.0 (6.0) 10.1(6.1) 6.8 (4.6) 7.3 (5.0) CO (L/min) .45 (.42) .62 (.50) .41 (.33) .46(.42) TPR (pru) .06 (.06) .04 (.05) .06 (.05) .02(.06) Values for SBP, DBP, HR, CO, and TPR are unadjusted means (SD). HOST—hostility; EX—interpersonal emotional expression; SBP—systolic blood pressure; DBP—diastolic blood pressure; HR—heart rate; CO—cardiac output; TPR—total peripheral resistance. Table options Analyses for both HR, F(1,90) = 5.06, p < .05, and SBP, F(1,88) = 4.46, p < .05, yielded significant main effects of hostility group that were qualified by significant Hostility Group × Task interactions [for HR: F(3,270) = 2.86, p < .05; for SBP: F(3,264) = 2.78, p < .05]. Follow-up analyses for speech preparation yielded significant main effects of hostility group for both HR (MHigh-Hostile = 4.3, SD = 5.4; MLow-Hostile = 8.9, SD = 8.0), F(1,90) = 9.59, p < .01, and SBP (MHigh-Hostile = 8.3, SD = 6.3; MLow-Hostile = 11.0, SD = 8.9), F(1,93) = 8.74, p < .01, with high-hostile individuals exhibiting less reactivity than low-hostile individuals. Inspection of means suggested that high-hostile expressors were the least responsive group in terms of SBP, DBP, and TPR responses (see Table 2). While Hostility Group × Interpersonal Expression Group interactions were observed for both SBP, F(1,88) = 4.08, p < .05, and DBP, F(1,88) = 5.01, p < .05, follow-up analyses revealed only a significant main effect of expression group for high-hostile individuals, for DBP, F(1,43) = 12.86, p = .001, with high-hostile expressors being less reactive across all tasks than high-hostile suppressors (see Fig. 1). Mean (±SEM) diastolic blood pressure reactivity change scores, averaged across ... Fig. 1. Mean (± SEM) diastolic blood pressure reactivity change scores, averaged across tasks, for hostility by interpersonal emotional expression groups. Values are unadjusted means. HOST: hostility; EX: interpersonal emotional expression. Significant differences (⁎p < .01) are indicated. Figure options Although no Hostility Group × Interpersonal Expression Group interaction emerged for TPR, F(1,89) = 1.37, p > .10, the mixed ANCOVA did result in a main effect of interpersonal expression group, F(1,89) = 5.24, p < .05, qualified by an Interpersonal Expression Group × Gender interaction, F(1,89) = 4.56, p < .05, as well as a Hostility Group × Gender interaction, F(1,89) = 6.64, p = .012. Follow-up analyses showed that male expressors exhibited lower TPR responses than male suppressors, F(1,35) = 9.80, p < .01 (see Fig. 2). No significant hostility group differences were observed for men, F(1,35) = 3.11, p > .05, or women, F(1,35) = 1.45, p > .10. Mean (±SEM) total peripheral resistance reactivity change scores, averaged ... Fig. 2. Mean (± SEM) total peripheral resistance reactivity change scores, averaged across tasks, for gender by interpersonal emotional expression groups. Values are unadjusted means. EX: interpersonal emotional expression. Significant differences (⁎p < .01) are indicated. Figure options 3.3. Hostility × interpersonal emotional expression group differences in cognitive appraisals Mixed ANCOVAs performed on cognitive appraisal ratios yielded significant main effects of both hostility group, F(1,90) = 5.00, p = .016, and expression group, F(1,90) = 4.26, p < .05. Relative to their low-hostile counterparts (M = .60, SD = .26), high-hostile individuals (M = .72, SD = .28) reported greater overall threat (and lower overall challenge) appraisals of all tasks. Conversely, emotional expressors (M = .61, SD = .27) reported lower overall threat (and greater challenge) appraisals than suppressors (M = .72, SD = .28). 3.4. Hostility × interpersonal emotional expression group differences in state affect Results for negative state affect showed only a hostility group main effect, F(1,90) = 4.28, p < .05. Specifically, high-hostile individuals reported higher negative affect than low-hostile participants at baseline and maintained this elevated negative emotion across the tasks (MHigh-Hostile = 18.3, SD = 4.83; MLow-Hostile = 16.1, SD = 3.71). Results for positive affect revealed a Hostility Group × Interpersonal Expression Group interaction, F(1,90) = 7.79, p < .01. Follow-up analyses indicated that high-hostile expressors reported lower levels of positive affect at baseline and during the tasks than high-hostile suppressors, F(1,44) = 8.79, p < .01 (see Fig. 3). Mean (±SEM) positive affect (PA) levels, averaged across baseline and tasks, for ... Fig. 3. Mean (± SEM) positive affect (PA) levels, averaged across baseline and tasks, for hostility by interpersonal emotional expression groups. Values are unadjusted means. HOST: hostility; EX: interpersonal emotional expression. Significant differences (⁎p < .01) are indicated.