Assessing the antecedents and consequences of threat appraisal of an acute psychosocial stressor: the role of optimism, displacement behavior, and physiological responses

Abstract

The feeling of stress is increasing in today’s societies, particularly in young adults subjected to social evaluative situations in highly competitive academic and work contexts. Threat appraisal is a primary and fundamental reaction when people face a stressful situation. The aim of this study was to investigate the role of dispositional optimism as an antecedent and displacement behavior as a consequence of threat appraisal of a social-evaluative situation of stress. A second objective was to verify the moderating role of physiological responses to stress (heart rate and cortisol reactivity) in the relationship between threat appraisal and displacement behavior. To do this, we combined the Trier Social Stress Test (TSST) with ethological analysis, self-report questionnaires, and physiological data. As expected, people who scored higher on dispositional optimism perceived stress as less threatening, and a higher perception of threat was positively related to displacement behavior patterns. Moreover, the results showed that threat appraisal fully mediates the relationship between dispositional optimism and displacement behavior, and that only heart rate reactivity (not cortisol) moderates the relationship between threat appraisal and displacement behavior.

Keywords:

• Dispositional optimism
• threat appraisal
• heart rate
• cortisol
• displacement behavior

Introduction

When facing a stressful situation, different responses might be considered, such as the stress appraisal and the physiological and behavioral responses (Blascovich & Tomaka, 1996). Moreover, it might be important to understand which personality traits influence stress responses. Particularly, optimism has attracted attention due to its potential protective influence. It is defined as an expectation that more good and desirable things will happen to us in the future than bad things (Scheier & Carver, 1985). The fact that dispositional optimism may have important implications for the way people cope and behave when faced with daily stress was described by Carver and Scheier’s (1981) self-regulation model. Scheier and Carver’s (1987) work suggests that when faced with a potentially challenging situation, optimists use more effective coping strategies that reduce the potential negative effects on behavioral performance.

When coping with stress, individuals assess both the situational demands (i.e. uncertainty involved in a performance situation) and their own coping resources to overcome them. Whenever people perceive they do not have the necessary coping strategies to overcome a stressful situation, they evaluate it as a threat (Gaab, Rohleder, Nater, & Ehlert, 2005). Threat appraisal is a negative stress-related mental state where a person anticipates personal harm due to challenges or novel situations (Feldman, Cohen, Hamrick, & Lepore, 2004). Due to optimistic individuals’ belief that they will be able to cope successfully with challenging situations, they will tend to perceive a lower threat appraisal.

Mental states produced by a stressful situation influence the way people behave (Blascovich & Tomaka, 1996). Some studies show that prior to or during a public speaking task, people with negative states, such as negative mood and/or anxiety, display more displacement behaviors (DB) (Troisi et al., 1996; Villada et al., 2014). DB are defined as stress-related movements focused on one’s body, such as grooming behavior (i.e. head scratching, licking and biting lips, etc.), iterative movements, and manipulation of objects (twisting and fiddling with fingers, twisting rings, etc.) (Troisi, 2002).

The relationship between mental states and DB might also be influenced by physiological responses to stress. Some studies have observed a negative relationship between cardiac (Pico-Alfonso et al., 2007; Villada et al., 2014) and cortisol (Pico-Alfonso et al., 2007; Villada et al., 2014) responses to stress and DB, however, the underlying moderated physiological mechanisms between some mental states (i.e. threat appraisal) and DB has not yet been studied.

The aim of this study was to investigate the role of dispositional optimism as an antecedent and DB as a consequence of threat appraisal of an acute psychosocial stressor, and verify the moderating role of heart rate (HR) and cortisol in the relationship between threat appraisal and DB. We expected that: (H1) dispositional optimism would negatively predict threat appraisal; (H2) threat appraisal would positively predict DB; and (H3) threat appraisal would fully mediate the relationship between dispositional optimism and DB. Finally, considering threat appraisal as an antecedent of DB, it was hypothesized that the physiological stress response, HR (H4a) and cortisol (H4b), would negatively moderate the relationship between the threat appraisal and DB. Due to the sex differences reported in the psychophysiological responses to stress, sex was taken into account in the corresponding statistical analyses (Kudielka, Buske-Kirschbaum, Hellhammer, & Kirschbaum, 2004; Liu et al., 2017; Reschke-Hernández, Okerstrom, Bowles Edwards, & Tranel, 2017; Wilhelm et al., 2017).

Methods

Participants

The final sample was composed of 82 healthy young adults (37 men and 45 women) (mean ± SEM: Age = 24.98 ± 0.55 years old). No sex differences were observed for age (from 20 to 40 years old) (men: M = 25.81, SEM ± 0.9; women: M = 24.31, SEM ± 0.6; F(1,80) = 1.831, p = .18) or SES (SES: ranging from 1 = low SES to 10 = high SES, Adler, Epel, Castellazzo, & Ickovics, 2000) (men: M = 5.89, SEM ± 0.1; women: M = 6.28, SEM ± 0.1; F(1,80) = 2.540, p = .11), but men had a higher body mass index (BMI = kg/m²) than women (men: M = 25.13, SEM ± 0.6; women: M = 22.27, SEM ± 0.4, F(1,80) = 14.265, p = .001). They were graduate and postgraduate students in a wide range of majors at the University of Valencia. All participants were volunteers, and at the end of the experimental session they received feedback from an expert interviewer about how to improve their individual performance on a job interview.

The following inclusion criteria were examined through self-reports: (i) Spanish nationality; (ii) age between 20 and 40 years; (iii) educational level between Secondary school and postgraduate studies; (iv) not smoking more than five cigarettes per day; (v) no alcohol or any other drugs; (vi) no visual or hearing problems; (vii) no cardiovascular, endocrine, neurological, or psychiatric diseases; (viii) not having been under general anesthesia once or more than once in the past year; (ix) not having experienced a major stressful life event during the past year; (x) not using any medication directly related to cardiac, emotional, or cognitive function, one that was able to influence hormonal levels, such as glucocorticoids or β-blockers, antidepressants, benzodiazepines, asthma medication, thyroid therapies, or psychotropic substances. Volunteers were assessed to verify that they met the experiment’s inclusion criteria. Subjects who fulfilled the criteria were asked to attend sessions that took place in a laboratory at the Faculty of Psychology.

Before each individual session, participants were asked to maintain their general habits, sleep as much as usual, refrain from heavy physical activity the day before the session, and not consume alcohol since the night before the session. Additionally, they were instructed to drink only water, and not eat, smoke, or take any stimulants, such as coffee, cola, caffeine, tea, or chocolate, 2 h prior to the session.

The study was conducted in accordance with the Declaration of Helsinki, and the protocol and conduct were approved by the Ethics Research Committee of the University of Valencia. Upon arrival at the laboratory, all the participants received verbal and written information about the study and signed an informed consent form.

Procedure

The study involved an individual session that lasted approximately 90 min. Based on the circadian rhythm of cortisol, all the sessions took place between 16.00 and 19.00 h, when the cortisol level is lower than in the morning (Kudielka, Schommer, Hellhammer, & Kirschbaum, 2004). In normal individuals, cortisol reaches its peak level during the morning and slowly declines throughout the day (Chan & Debono, 2010). The experimental sessions were composed of different phases (Figure 1). Upon arrival at the laboratory, the experimenter verified that participants had followed the instructions given previously.

Figure 1. Timeline of the TSST. Dotted lines depict the time of HR register. Salivary cortisol samples = 1, 2, 3, and 4 °Co. Dotted lines represent heart rate variability samples. LOT-R: life orientation test-revised.

To produce stress, we subjected the participants to a modified version of the Trier Social Stress Test (TSST, Kirschbaum, Pirke, & Hellhammer, 1993). This test includes a highly competitive component where the subject has to persuade the committee that she/he is “the best applicant” for the position (Salvador & Costa, 2009). The modifications were: (i) all the phases of the TSST took place in the same room and (ii) the committee was composed of only one person who had been introduced as an expert in human resources. The session started with a 40 min habituation phase. During that time, participants had to fill out a general questionnaire related to demographics and anthropometric data and the Life Orientation Test-Revised (LOT-R, Scheier, Carver, & Bridges, 1994); for the last 10 min, they were left alone to rest mentally and physically. Next, during the introduction phase, participants were told about the job interview task. Immediately after that, they had to fill out the Primary Appraisal Secondary Appraisal scale (PASA) (Gaab et al., 2005). Before the beginning of the TSST, participants had 5 min to prepare their presentations. During this preparation phase, individuals had to write down their main ideas about what to say during the job interview. However, they could not use these notes during the speech task. The TSST protocol consisted of 5 min of free speech (job interview), followed by a 5 min arithmetic task. In the job interview, the participant’s main goal was to convince the interviewer that she/he was the best candidate for his/her “dream job”. The participants stood at a distance of 1.5 m from the evaluator. In addition, a video camera, a microphone, and a monitor where subjects could see their performance were clearly visible. Both the speech and arithmetic tasks were videotaped and subsequently, the DB was assessed.

Measures

Displacement behavior

Three observers (two men and one woman) were trained to interpret and reliably rate the participants’ behavior from the video recordings on the basis of a modified version of the Ethological Coding System for Interviews (ECSI) (Troisi, 2002). We analyzed two main behavioral categories: (1) Hand movement DB (head scratching, twisting and fiddling with fingers, touching and twisting rings, constant arrangement of clothing, covering their mouth with their hands and touching their lips, and clicking their fingers); (2) Body movement DB (repeated rocking forward and backward, licking and biting lips, and repeated arm movements. Each DB was assessed as present = 1 or not present = 0, with a frequency of 20. Each video lasted 5 min (job interview). The total score indicates the frequency of DB in each category. The observers rated the full set of videos on two separate occasions, with the second rating taking place three weeks after the first rating. To assess the repeatability of the ratings from a given observer, we first calculated an intra-observer correlation coefficient (ICC) for each item, based on the two separate ratings nested within each subject. Then, we averaged the observers’ first and second ratings for each item, and we used these average scores to assess the degree of inter-observer ICC among the three observers. The repeatability of the ratings of intra- and inter-observers’ ICC ranged from 0.78 to 0.95 (0.84 ± 0.15) for the three observers.

Life orientation test-revised (LOT-R; Scheier et al., 1994)

LOT-R is composed of 10 items answered on a 5-point Likert scale ranging from 0 (strongly disagree) to 4 (strongly agree). Three items measure optimism (e.g. “In uncertain times, I usually expect the best”), and three items measure pessimism (e.g. “If something can go wrong for me, it will”); the remaining items are distractors. Pessimism items were reversed to obtain a one-dimensional measure of dispositional optimism. We employed the Spanish version (Otero, Luengo, Romero, Gómez, & Castro, 1998), which has shown adequate reliability (α = 0.78) (Ferrando, Chico, & Tous, 2002).

Threat appraisal

Threat appraisal was evaluated using a situation-specific subscale of the PASA (Gaab et al., 2005). This subscale was employed to assess threat appraisal processes before performing the TSST, based on transactional stress theory (Lazarus & Folkman, 1984). The subscale has four items, rated on a 6-point Likert scale ranging from 1 (strongly disagree) to 6 (strongly agree) (e.g. I do not feel threatened by the situation). The scale was translated into Spanish and back-translated. In our sample, Cronbach’s alpha for the scale was 0.76. The PASA scale was administered at the end of the introductory phase of the TSST.

Heart rate

HR data were continuously recorded during the entire session using a Polar^©RS800cx watch (Polar CIC, Bethpage, NY), which consists of a chest belt for detection and transmission of heartbeats and a Polar watch for data collection and storage. The transmitter is located on the chest belt, which is placed on the solar plexus and transmits HR information to the receiver (Polar watch). The data collected by the Polar watch were downloaded, stored in the Polar ProTrainer5TM program in the computer, and analyzed using HRV Kubios Analysis software (Biomedical Signal Analysis Group, University of Kuopio, Kuopio, Finland). Following the recommendations of the Task Force (1996), we analyzed HR in periods of 5 min. Whereas the job interview and arithmetic task phases lasted 5 min each, the habituation, preparation, and recovery phases lasted longer than 5 min; for this reason, we chose the central 5 min of each phase. HR analysis failed to detect the HR measurements in two women and three men; therefore, these subjects were excluded from the HR statistical analyses.

Cortisol

We measured the activity of the hypothalamic-pituitary-adrenal (HPA) axis by analyzing the cortisol from the five salivary samples taken throughout the session (see Figure 1). Participants provided saliva samples by using salivettes (Sarstedt, Nümbrecht, Germany). They were instructed to keep the cotton swab in their mouths for exactly 2 min, not chew the cotton, and move the swab around in a circular pattern to collect saliva from all salivary glands. The samples were centrifuged at 3000 rpm for 5 min, resulting in a clear supernatant of low viscosity that was stored at −80 °C until the analyses were performed in the Central Research Unit (Unidad Central de Investigación) of the Faculty of Medicine, University of Valencia (Spain). The salivary cortisol samples were analyzed by a competitive solid phase radioimmunoassay (tube coated), using the commercial kit Spectria Cortisol RIA (cat. Nu 06119) from Orion Diagnostica (Espoo, Finland). Assay sensitivity was 0.8 nmol/L, and within-assay and inter-assay variation coefficients were all below 8%.

Statistical analysis

HR and cortisol values were logarithmic transformed because they did not have a normal distribution after the Kolmogorov–Smirnov and Levene tests were applied. Delta changes (Δ) for HR and cortisol were calculated by subtracting baseline levels (HR: habituation; cortisol: −15 min) from the highest indexes (HR: job interview; cortisol: +20 min). Three participants in the cortisol data (one female and two males) and one participant in the HR data (one female) were removed from the analyses because their indexes differed by more than 3 SD from the total sample mean.

Because sex might affect the psychological and physiological responses to TSST (i.e. Buchanan et al., 2010; Liu et al., 2017; Reschke-Hernández et al., 2017; Stephens, Mahon, McCaul, & Wand, 2016), sex differences in HR and cortisol were assessed using analysis of variance (ANOVA) for repeated measures, with sex (men vs. women) as between-subject factor and time (cortisol: −15, + 5, + 10, + 20, + 45 min; HR: habituation, preparation, job interview, arithmetic task, and recovery) as a within-subject factor. Moreover, we analyzed possible sex differences in dispositional optimism, threat appraisal, displacement behavior, as well as demographic and anthropometric measures using multivariate analysis of variance (MANOVA).

Following Preacher and Hayes (2004), first a mediation procedure based on nonparametric resampling, known as bias-corrected bootstrapping, was conducted to assess the mediating effect of threat appraisal in the relationship between dispositional optimism and DB (PROCESS model number 4). Dispositional optimism was entered as the independent variable, threat appraisal was entered as the mediator variable, and DB was entered as the dependent variable. Then, we performed moderated mediation analyses to estimate the moderator effect of ΔHR and Δcortisol in the relationship between threat appraisal and DB, with threat appraisal as a mediator variable between dispositional optimism and DB (PROCESS model number 14). In this case, we added ΔHR and Δcortisol as moderating variables in the second stage of the previous paths. Bootstrapping makes it possible to gather many alternative versions of a single statistic that is usually only calculated from one sample. Bootstrap data resampling procedures establish confidence intervals (CIs) for testing the statistical significance of an indirect effect (Shrout & Bolger, 2002). The analysis was based on 10,000 bootstrap iterations, and the CI was set to 95%, as recommended by Mallinckrodt, Abraham, Wei, and Russell (2006). Due to sex differences in physiological and BMI responses found, sex and BMI were included as covariate variable in the mediation and moderated mediation analyses with ΔHR and Δcortisol.

All statistical analyses were carried out using SPSS version 20 (SPSS Inc., Chicago, IL). If not explicitly indicated, data are presented as mean ± SEM. For an easier interpretation of the figures, the values are presented in raw values and not in logarithmic transformed values.

Results

Preliminary analyses

For cortisol, ANOVA showed a main effect of sex (F(1,75) = 21.416, p = .01); in this case, men had higher cortisol than women. Moreover, the main effect of time (F(1.852,138.931) = 5.572, p = .006) was also significant. Cortisol concentrations increased immediately after the job interview (−15 min sample vs. +5 min sample, p = .01), and they continued to increase until reaching peak levels 20 min after the onset of the stress task (−15 min sample vs. +20 min sample, p = .007). Afterwards, cortisol levels decreased, reaching baseline levels in the last saliva sample (−15 min sample vs. +45 min sample, p = .84). The interaction between time and sex was not statistically significant (p = .06) (Figure 2).

Figure 2. Salivary cortisol concentrations (i) and heart rate (ii) for men and women during TSST. The repeated measures ANOVA showed the main effect of time on both salivary cortisol (*p = .006) and HR (**p < .001). Moreover, it indicates the main effect of sex, where men generally had higher cortisol concentrations than women (*p = .01); and women generally had higher HR than men (*p = .01). Although the interaction between time and sex was not statistically significant for salivary cortisol (p = .06) or HR (p = .20) in this study, we carried out planned comparisons that showed sex differences in all the phases of the protocol for cortisol (p  < .02). However, there was a significant difference in habituation for HR (p  = .02). Depicted values are means and error bars represent the SEM. *p  < .05; **p  < .005.

For HR, the ANOVA showed a main effect of sex (F(1,74) = 6.341, p = .01); overall, women had a higher HR than men. The main effect of time (F(3.084,228.212) = 130.390, p < .001) was also significant. HR increased immediately after the habituation phase (habituation vs. preparation phase, p < .001), and it continued to increase until reaching its peak rate during the job interview phase (habituation vs. job interview phase, p < .001). Then, HR decreased (habituation vs. recovery phase, p = .09). The interaction between time and sex was not statistically significant (p = .20) (see Figure 2).

No sex differences were found for threat appraisal (men: M = 4.52, SEM ± 0.1; women: M = 4.33, SEM ± 0.1; F(1,80) = 0.238, p = .20), dispositional optimism (men: M = 22.56, SEM ± 4.3; women: M = 22.13, SEM ± 3.2; F(1,80) = 3.829, p = .60), and DB (men: M = 6.98, SEM ± 2.1; women: M = 6.77, SEM ± 1.4; F(1,80) = 0.921, p = .73).

Testing mediation model

Dispositional optimism did not have a significant effect on DB (p = .09). Moreover, even though the B coefficient found was small (B = −0.05), dispositional optimism has a significant negative effect on threat appraisal (p = .03). At the same time, threat appraisal has a significant positive effect on DB (p = .04) (Figure 3).

Figure 3. Moderation and mediation analysis with ΔHR and Δcortisol, using bias-corrected bootstrapping in conjunction with multiple regression analysis. Solid lines represent a significant direct effect; dashed lines indicate non-significant effects. Numbers on the lines show B and p values. This figure indicates that dispositional optimism was negatively related to threat appraisal (B = −0.05, SE = 0.02, t = −2.22, p = .03), and threat appraisal was positively related to DB (B = 0.83, SE = 0.40, t = 2.06, p = .04). Dispositional optimism did not have a significant effect on DB (B = −1.13, SE = 0.08, t = −1.68, p = .09). Moreover, it indicates that ΔHR moderated the relationship between threat appraisal and DB (B = −19.09, SE = 8.91, t = −2.14, p = .03), whereas Δcortisol did not (B = −2.68, SE = 2.98, t = −0.89, p = .37). *p < .05.

The bootstrapping procedure (Shrout & Bolger, 2002) was performed to verify the significance of the full effect of threat appraisal on the relationship between dispositional optimism and DB. The conditional indirect effect of dispositional optimism on DB through threat appraisal was significant (B = −0.04, SE = 0.02, 95% CI = −0.10 to −0.006) because the 95% CI does not include zero.

Testing moderated mediation model with ΔHR

The interaction effect between threat appraisal and ΔHR on DB was negative and statistically significant (p = .03) (Figure 3). In order to make the interpretation of significant interaction effects easier, a graphic representation was elaborated to plot threat appraisal against DB for low and high levels of ΔHR (1 SD below the mean and 1 SD above the mean, respectively). The values for the regression line slopes (Aiken, West, & Reno, 1991) demonstrate that the relationship between threat appraisal and DB was positive and significant under low levels of ΔHR. However, the relationship between threat appraisal and DB was not statistically significant under high levels of ΔHR (Figure 4).

Figure 4. Displacement behavior during the TSST as a function of threat appraisal and ΔHR. Low threat appraisal/ΔHR was defined as 1 SD below the mean; high threat appraisal/ΔHR was defined as 1 SD above the mean. The t-tests revealed that individuals with low ΔHR and high threat appraisal manifested more DB than individuals with low ΔHR and low threat appraisal (*p < .05). Finally, the relationship between threat appraisal and DB was not statistically significant under high levels of ΔHR (p > .05). *p < .05.

Finally, the conditional indirect effect of dispositional optimism on DB through threat appraisal was examined for three values of ΔHR: 1 SD below the mean, the mean, and 1 SD above the mean. The findings showed that ΔHR enhanced the indirect effect of dispositional optimism on DB through threat appraisal when ΔHR was low (B = −0.08, 95% CI = −0.19 to −0.016) and moderate (B = −0.04, 95% CI = −0.10 to −0.004), but not if it was high (B = 0.002, 95% CI = −0.04 to 0.07), a result that is not statistically significant because the 95% CI includes zero.

Testing moderated mediation analysis with Δcortisol

Figure 3 depicts the results of the moderated mediation analysis for Δcortisol. It shows that the interaction terms between threat appraisal and Δcortisol on DB were not statistically significant (p = .37).

Discussion

The main aim of this study was to verify the antecedent role of dispositional optimism and the consequent role of threat appraisal of an acute psychosocial stressor on DB, and investigate the moderating role of heart rate and cortisol in the relation between threat appraisal and DB. As expected, our results show that people with higher dispositional optimism assessed the stressful task as less threatening. This result agrees with Scheier and Carver’s (1987) theory indicating that optimistic people trust their coping strategies in facing challenging situations in everyday life. Moreover, it has been reported that optimistic individuals have coping strategies related to the elimination, reduction, or control of the stressors (Nes & Segerstrom, 2006). Thus, people who believe they have the coping strategies to overcome a challenge perceive it as less threatening (Lazarus & Folkman, 1984). Considering that in typical daily life frequent perception of threat is maladaptive and associate with a dysregulation in hippocampal circuits, endocrine, and autonomic output, as well as with a cognitive and general health decline (Thayer, Åhs, Fredrikson, Sollers, & Wager, 2012), this result supports the idea of a health protection role of optimism in stressful situations (Carver, Scheier, & Segerstrom, 2010; Scheier & Carver, 1993). In fact, some studies have observed that people with high dispositional optimism manifest better physical and mental health, even in pathological situations such as type 2 diabetes, where the stress perception is high (Puig-Pérez, Hackett, Salvador, & Steptoe, 2017; Steptoe et al., 2014).

In our study, threat appraisal showed a positive relationship with DB. When the task was appraised as more threatening, more DB was displayed. In addition, mediation analysis indicated that threat appraisal fully mediated the relationship between dispositional optimism and DB. No direct relationship between optimism and DB was found. To the best of our knowledge, our study is the first one to take DB into consideration as a consequence of threat appraisal. However, previous literature supports this finding, suggesting that some other negative mental states (i.e. depression and anxiety) might influence the level of DB during stressful tasks (Shreve, Harrigan, Kues, & Kagas, 1988; Troisi et al., 1996; Troisi et al. 2000; Troisi, Spalletta, & Pasini, 1998; Villada et al., 2014), similarly to grooming behavior in animal research (Troisi, 2002). Hence, we considered that threat appraisal, as a temporary negative mental state, might influence the level of DB during a social evaluative stress situation. Specifically, according to these authors, DB is manifested to reduce stress when high psychological activation (i.e. anxiety) exists (Troisi, 2002; Villada et al., 2014). Thus, threat appraisal, as psychological activation in response to stress, leads to DB. However, because DB requires psychological activation and dispositional optimism is not a stressful state of activation, but rather a positive personality trait, DB is not directly influenced by it. Hence, as discussed above, dispositional optimism is related to threat appraisal, which, in turn, would affect the use of DB to reduce psychological activation produced by stress.

Interestingly, our results also show that HR reactivity negatively moderates the relationship between threat appraisal and DB. A similar result was not found for cortisol reactivity. Regarding HR, the relationship between threat appraisal and DB is positive when low and medium HR reactivity exists. However, the relationship between threat appraisal and DB is not statistically significant when HR reactivity is high. In other words, when HR reactivity is low/medium, people who perceive low threat manifest low DB, and people who perceive high threat manifest high DB. As mentioned above, DB might regulate the individual level of arousal created by a high level of threat appraisal, even without greater sympathetic activation. Thus, high/medium levels of threat appraisal lead to DB, even when cardiovascular activation is low. This might be the case of one of the three types of psychophysiological responses to cognitive tasks described by Boudarene, Legros, and Timsit-Berthier (2002). They indicate that “biological silence” is a psychophysiological reaction to a stressful task where subjects manifest high emotional responses without increases in physiological responses. Thus, a “biological silence” reaction might lead people to use DB to rebalance their psychological state and reduce the response to stress. According to Troisi (2002), DB are likely to be behavioral elements of the adaptive psychophysiological stress response, possibly having an anxiolytic effect; this is also found in the grooming behavior that has been involved in de-arousal from stressful situations (see Spruijt, van Hooff, & Gispen, 1992). Specifically, threat appraisal appears to be a fundamental emotional state associated with low or medium HR reactivity to the use of DB. Furthermore, we did not find a significant moderating role of high HR reactivity in the relationship between threat appraisal and DB, and the anticipation of stress might play an important role in explaining this result. Indeed, DB may make it possible to reduce the cardiac hyper-activation that appears prior to the stress task (anticipation). High anticipation results in low cardiac reactivity, whereas low anticipation results in high cardiac reactivity (cardiac reactivity is calculated by subtracting the higher HR level from the baseline). This finding agrees with some studies that observed a positive relationship between an anticipatory negative mental state and DB (Troisi et al. 1996; Villada et al., 2014). Moreover, our result demonstrates that the anticipation of threat appraisal has a positive relationship with DB. Thus, lower/medium cardiac reactivity due to higher HR anticipation strengthens the relationship between anticipatory threat appraisal and DB. By contrast, higher HR reactivity due to lower HR anticipation does not affect this relationship.

Regarding cortisol responses, we did not find a moderating role in the relationship between threat appraisal and DB. Therefore, cortisol does not strengthen or weaken the link between threat appraisal and DB. Taking into account that cortisol increases in this study could be considered moderate, these behavioral patterns may not be necessary to regulate HPA axis activity. This result agrees with Campbell and Ehlert (2012), who reviewed a total of 49 studies in which the paradigm of social stress used was always the TSST. They concluded that stress appraisal and physiological reactions are not always correlated. That is, cortisol does not seem to reflect a greater experience of stress or anxiety, which seems to be fundamental in the display of DB. Moreover, as Villada, Hidalgo, Almela, and Salvador (2016) stated, a moderate cortisol increase might have a positive function, reflecting the response of preparing to deal with stress.

This work adds knowledge to the study of DB as part of the acute stress response. It reveals the important antecedent role of some personality traits, such as dispositional optimism, in triggering DB through the threat appraisal of stress. Moreover, to the best of our knowledge, it is the first study to observe the moderating role of HR reactivity in the relationship between threat appraisal and DB. Moreover, it provides insight into the role of DB in communicating information about the individual’s emotional state and personality traits. Indeed, DB is believed to reflect the state of tension brought about by the social context (Maestripieri, Schino, Aureli, & Troisi, 1992), providing more accurate information about the subject’s emotional state than verbal statements and facial expressions (Troisi, 2002). In agreement with Mohiyeddini and Semple (2013), DB might be considered “cues” rather than “signals,” which might be quite useful during various evaluative situations such as clinical or job interviews. In fact, DB might help clinicians or psychologists to detect people’s personality traits and coping strategies, or help recruiters to detect candidates’ stress management abilities, which would be beneficial for the company’s hiring process.

Finally, whether and how threat appraisal mediates the relationship between different personality traits and other stress-related behaviors remains poorly understood. Further studies in this area are required to provide new insight into the relationships among threat appraisal, personality traits, and behavioral stress responses in interpersonal communication. Moreover, due to the nature of the study (experimental design and protocol), its internal validity is high, but its external validity is limited.

In conclusion, our study greatly contributes to better understanding the relationships among threat appraisal, dispositional optimism, and DB. Indeed, it has demonstrated the fundamental mediating role of threat appraisal in the relationship between optimism and DB. Moreover, a moderating role of HR reactivity was found in the second step of this mediation, namely in the relationship between threat appraisal and DB. Nevertheless, further studies are needed, taking natural settings into consideration.

Acknowledgments

We are grateful to Ms. Marta Garcia-Lluch for her collaboration during the experimental process, Prof. Mario Martinez-Córcoles for his support in the statistical analyses, and Ms. Cindy DePoy for the revision of the English text.

Disclosure statement

The authors state that there are no conflicts of interest associated with this research.

Additional information

Funding

This research was supported by the Spanish Education and Science Ministry [PSI2013/46889, PSI2016-78763] and Generalitat Valenciana [GRISOLIAP/2011/082, PROMETEOII2015/020].