خوی بازدارنده و بی بندوباری و واکنش پذیری استرس اتونوم

Abstract We examined the relationship of temperament dimensions serving as markers for Gray’s behavioral activation system (BAS) and behavioral inhibition system (BIS) with autonomic stress reactivity in 35 middle-aged men. Temperament was measured using the Strelau Temperament Inventory — Revised. Skin conductance responses and inter-beat interval were measured during administration of the Rorschach test. The results showed that temperamental activation was positively related to the task-level of and task-induced change in respiratory sinus arrhythmia (RSA) amplitude, but unrelated to heart rate (HR) reactivity. Temperamental inhibition was negatively associated with the task-level of electrodermal activity and task-induced change in RSA amplitude, and positively associated with HR reactivity. The findings are in part contrary to the hypotheses presented in the literature. They also suggest that the temperamental inhibition–HR reactivity relationship is mediated by the parasympathetic nervous system.

Introduction Exposure to acute stressors is an ubiquitous part of everyday life. Consequently, all research providing information on factors underlying stress vulnerability is of importance. Especially important are the factors which explain inter-individual differences in stress reactivity because they may increase our understanding of why coping with daily stressors has greater health consequences for some individuals than for others. Individual variation in autonomic nervous system (ANS) reactivity to brief psychological stressors might explain why daily annoyances are more likely to call forth somatic endpoints in some individuals than in others (e.g. Pollak, 1991, Matthews et al., 1992 and Cacioppo, 1994). To increase our understanding of stress-related individual differences in cardiovascular and neuroendocrine reactivity, recent research has discussed the concept of temperament (e.g. Calkims and Fox, 1995, Gunnar, 1995 and Rothbart et al., 1995). Temperament consists of biologically rooted individual differences in behavior tendencies that are present early in life and are relatively stable across time and situations (Bates, 1989); temperament highlights a person’s behavioral style, the how of behavior. Given that individual differences in temperament are attributable to differences in neural and physiological functions (e.g. Rothbart, 1989, Steinmetz, 1995 and Strelau, 1995), it has been suggested that biologically rooted temperament may operate as a stress-buffer or increase stress-vulnerability via temperament-related cardiovascular and neuroendocrine functions (e.g. Gunnar, 1995). Researchers sharing the suggestion that reactivity to stressors is an important aspect of temperament frequently refer to Gray’s neurophysiological model of temperament (Gray, 1979 and Gray, 1982), which organizes a large number of findings that would otherwise appear unrelated. According to Gray (1991) temperament reflects individual differences in predispositions towards particular kinds of emotion. His model posits that behavioral reinforcement is mediated by three fundamental motivational systems, which involve separate sets of interacting brain structures. The systems are the behavioral inhibition system (BIS), the behavioral activation system (BAS), and the fight–flight system (FFS). It has been suggested that the dispositional balance of these systems to respond to external cues underlies individual differences in temperament (e.g. Rothbart, 1989). The BIS is responsible for the inhibition of behavior (i.e. passive avoidance and extinction) and it responds to negatively conditioned stimuli (i.e. punishment and non-reward) or novel stimuli. It is a substrate for anxiety, and is linked to an increase in arousal and attention. The BAS mediates responses for positively conditioned stimuli (i.e. reward and non-punishment). It is associated with pleasurable emotions and non-specific arousal, and it mediates the cardiac–somatic coupling. The FFS responds to unconditioned aversive stimuli, and its activity leads to defensive aggression or escape behavior. The FFS is not discussed any further in this study, however. Gray did not extend his theory to the traditional psychophysiological measures but stated that the significance of autonomic activity for emotional behavior ‘... is opaque’ (Gray, 1973, p. 434; cf. Fowles, 1980). Later, Fowles, 1980, Fowles, 1982, Fowles, 1983 and Fowles, 1988 and Fowles et al. (1982) proposed a model of cardiovascular and electrodermal activity to be applied to Gray’s theory. According to Fowles, tonic increases in heart rate (HR) reflect central activity of the BAS, or more specifically, the incentive-related activation of the BAS, while activation of the BIS was presumed to lead to an increase in tonic electrodermal activity (EDA), particularly in the frequency of non-specific skin conductance responses (NSCRs). There are many convincing studies suggesting that tonic heart rate changes can be used as an index of the BAS in conditions that demand active coping with acute stress in order to avoid punishment or to gain reward (e.g. Scher et al., 1984, Wright et al., 1992 and Wright and Dismukes, 1995; see also Sosnowski et al., 1991), although studies that have failed to reveal the predicted physiological outcome also exist (Sosnowski et al., 1991 and Clements and Turpin, 1995). The relationship between EDA and the BIS is more equivocal, however. Despite some positive evidence (see Fowles, 1980 and Clements and Turpin, 1995), other studies have failed to link EDA to the activity of the BIS (e.g. Sosnowski et al., 1991). This might be attributable, in part, to the fact that EDA responds unspecifically to a wide range of stimuli (e.g. Fowles, 1980; see also Boucsein, 1992). Although empirical findings provide only limited support for the predictions made by Fowles’ ‘three-arousal model’ (Fowles, 1980), it provides an alternative to a simple arousal-based interpretation of autonomic activity and offers a solid theoretical base for research on stress psychology. The work of Kagan and co-workers is also of interest here (Kagan, 1989 and Kagan et al., 1989). In a longitudinal study, Kagan identified behaviorally inhibited children who were characterized by withdrawing, shy, and timid behavior, and distinguished uninhibited temperaments from inhibited temperaments, uninhibited children being sociable and talkative (Kagan et al., 1984). In addition to the qualitative differences at the behavioral level, the two temperaments have been found to differ in terms of ANS reactivity to novel stimuli. Inhibited temperament has been related to relatively high cardiac reactivity and low heart rate variability, as well as to other physiological indices of distress, including higher levels of morning cortisol (Kagan et al., 1987). It is of note that there appears to be some discrepancy between the predictions based on the models of Gray and Fowles on the one hand and the empirical findings by Kagan and co-workers on the other, although the operational definitions of Kagan’s and Gray’s constructs are conceptually consistent with regard to behavioral inhibition. According to the aforementioned models, temperamental inhibition is likely to produce a high level of EDA and low HR reactivity to a mild stressor, anxiety being an associated emotion. In contrast, Kagan and co-workers found that temperamental inhibition is primarily related to high cardiac reactivity. Thus, this issue requires further investigation. It is also important to notice that, in contrast to sympathetically mediated EDA, cardiac function involves both sympathetic and parasympathetic innervation. For example, cardiac acceleration (i.e. reactivity) may occur as a result of parasympathetic withdrawal, sympathetic activation, or both (Berntson et al., 1994). Some prior studies have shown that highly inhibited children tend to have sympathetically dominant responses to novel situations, whereas highly uninhibited children show parasympathetic dominance (Garcia-Coll et al., 1984 and Kagan et al., 1987). The differential relationship of temperamental inhibition with sympathetic and parasympathetic stress reactivity clearly deserves further inquiry, however. First, the aforementioned few studies have been carried out in children; there are no prior studies addressing this issue in adults. Second, the so-called inhibited vs. uninhibited child described by Kagan and co-workers is, in effect, a temperament category or cluster derived from a pattern of behavior. The relationship of individual temperament dimensions that may conceptually more closely correspond to the BIS with a specific pattern of sympathetic and parasympathetic reactivity has not been examined. Third, in the studies by Kagan and co-workers, the measures of HR variability used were the standard deviation of heart period (Kagan et al., 1984) and spectral power including non-respiratory frequencies (Kagan et al., 1987); these measures are not pure indices of parasympathetic function, but also reflect sympathetic influences on the heart. In contrast, spectral analysis allows a selective measurement of the so-called respiratory sinus arrhythmia (RSA), respiratory-locked oscillations in HR, which provides a non-invasive index of parasympathetic function. In view of the above considerations, the aim of the present study was to examine the relationship of temperamental dimensions related to the BAS and BIS with EDA and cardiac stress reactivity, specifically cardiac parasympathetic reactivity as measured by the RSA, during a mildly (or moderately) stressful psychological testing session (i.e. the Rorschach test; Exner, 1993) among healthy middle-aged men. According to the model of Gray and Fowles, it could be expected that: (a) the BIS and BAS are related to a decreased and increased amount of spontaneous verbal activity, respectively; (b) the BIS is related to an increased number of electrodermal responses; and (c) the BAS is related to increased HR. In addition, on the basis of Kagan’s research, it could be expected that behavioral inhibition is related to decreased RSA and increased HR.

Results 3.1. Psychological measures There were significant correlations between SE and SI and between SE and M (r=0.57 and 0.72, respectively, P<0.001), whereas the correlation between SI and M was not significant (r=0.19). On average, participants produced 21.3 verbal responses during the association phase of Rorschach testing (S.D.=10.0). Table 1 shows the correlations between the STI–R dimensions, initial mood, and number of verbal responses and the results of regression analyses with SE, SI, and M as the independent variables and initial mood and the number of verbal responses as the dependent variables. Table 1 indicates a clear pattern associating SI with increased energetic mood and calmness, and with decreased tension. There was also a significant positive correlation between M and tension. As expected, SE and SI were related to an increased and decreased number of verbal responses during the administration of the Rorschach test, respectively. Table 1. Correlations and partial correlations between the STI–R dimensions, initial mood, and behavioral activitya Correlations Standardized beta coefficientsb SE SI M SE SI M Mood Energetic 0.25 0.29* 0.22 −0.07 0.29 0.20 Tired −0.07 −0.12 −0.03 0.05 −0.18 0.01 Tension −0.05 −0.35** 0.27 −0.26 −0.27 0.48* Calmness 0.04 0.43*** −0.17 −0.15 0.57** −0.20 Behavioral activity Verbal responsesc 0.24 −0.08 0.16 0.52* −0.35* −0.15 a SE=Strength of excitation; SI=strength of inhibition; M=mobility; verbal responses=the number of verbal responses during the association phase of Rorschach testing. ***P<0.01, **P<0.05, *P<0.10. b SE, SI, and M have been entered together into the regression analysis as independent variables. c The duration of the task period is controlled for. Table options 3.2. Physiological measures Table 2 shows the means and standard deviations of the cardiac and behavioral measures during the baseline and task periods as well as EDA during the task; the number of NSCRs was near zero during the baseline period. On the average, participants exhibited a significant mean increase in HR relative to baseline values during the association phase of Rorschach testing (mean=2.31, S.D.=3.82, t=3.57, P=0.001). There was no overall task-induced change in RSA amplitude. Mean HR was negatively related to RSA amplitude during the baseline period (r=−0.64, P<0.001) and during the task period (r=−0.56, P=0.001). There was a non-significant negative correlation between baseline HR and task-induced HR change (r=−0.20) and a significant negative correlation between the baseline values of RSA amplitude and task-induced change in RSA amplitude (r=−0.43, P=0.01). Given these correlations, residualized change scores were applied in the further analyses of cardiac reactivity. Table 2. Cardiac, electrodermal, and behavioral measures during the baseline and task periodsa Measure Mean S.D. Baseline period HR 68.32 12.16 RSA amplitude 12.00 1.10 Task period HR 70.63 12.01 RSA amplitude 12.07 1.05 NSCR 1.43 1.09 Verbal responses 21.26 9.99 a HR=heart rate; RSA amplitude=the logarithm of spectral power within the 0.12–0.40-Hz range of heart rate variability; NSCR=the frequency of non-specific skin conductance responses/min; verbal responses=the number of produced verbal responses during the association phase of Rorschach testing. Table options Table 3 shows age-corrected partial correlations between the STI–R dimensions and cardiac and electrodermal measures during the baseline and task periods. SI was related to fewer NSCRs and greater HR reactivity during the association phase of Rorschach administration. In addition, M was related to a decreased RSA amplitude during the task period as well as greater task-induced reduction in the RSA amplitude. The relationships of task-induced change in NSCRs to other measures were not reported since the change in NSCRs was almost identical to the task-level of NSCRs. Table 3. Age-adjusted partial correlations between the STI–R dimensions and cardiac and electrodermal measuresa Physiological measure Ageb SE SI M Baseline period HR 0.13 0.06 −0.01 0.11 RSA amplitude −0.32* −0.07 0.09 −0.20 Task period HR 0.18 0.10 0.10 0.07 RSA amplitude −0.50*** −0.23 −0.04 −0.49*** NSCR −0.33** −0.17 −0.42** 0.06 Task-induced change HR 0.23 0.13 0.36** −0.12 RSA amplitude −0.39** −0.24 −0.14 −0.45*** a SE=strength of excitation; SI=strength of inhibition; M=mobility; HR=heart rate; RSA amplitude=the logarithm of spectral power within the 0.12–0.40-Hz range of heart rate variability; NSCR=the frequency of non-specific skin conductance responses/min. ***P<0.01, **P<0.05, *P<0.10. b Zero-order correlation. Table options The differential relationships of SE, SI, and M with each physiological measure were further examined using a series of regression analyses with SE, SI, and M as the independent variables and each physiological measure in turn as the dependent variable. These analyses were then repeated after including age as a covariate. Results of these analyses are presented in Table 4. Table 4. Results of multiple regression analyses (standardized beta coefficients): relationship of STI–R dimensions with cardiac and electrodermal measures before (Step 1) and after (Step 2) the inclusion of age in the regression modela Dependent variable Independent variables Step 1 Step 2 SE SI M Age SE SI M Baseline period HR −0.08 0.02 0.18 0.11 −0.01 −0.03 0.12 RSA amplitude 0.27 −0.05 −0.42 0.07 0.08 −0.25 −0.29 Task period HR −0.07 0.15 0.12 0.17 0.04 0.07 0.03 RSA amplitude 0.62** −0.31 −0.89*** −0.36** 0.38 −0.15 −0.69*** NSCR 0.13 −0.53** 0.01 −0.27 −0.05 −0.40* 0.16 Task-induced change HR 0.01 0.42** −0.16 0.21 0.15 0.32 −0.28 RSA amplitude 0.62** −0.39** −0.85*** −0.21 0.47 −0.29 −0.73*** a SE=strength of excitation; SI=strength of inhibition; M=mobility; HR=heart rate; RSA amplitude=the logarithm of spectral power within the 0.12–0.40-Hz range of heart rate variability; NSCR=the frequency of non-specific skin conductance responses/min. ***P<0.01, **P<0.05, *P<0.10. Table options The temperament dimensions were unrelated to the baseline measures of cardiac function. However, SE was positively associated with the task-level of and task-induced change in RSA amplitude; these associations did not remain significant after adjusting for age, however. SI was negatively associated with the task-level of NSCRs and task-induced change in RSA amplitude, and positively associated with task-induced change in HR; the relationships between SI and the task-induced changes in HR and RSA amplitude did not, however, remain significant after adjusting for age. In addition, M was negatively associated with the task-level of and task-induced change in RSA amplitude; adjustment for age did not change this. Fig. 1 and Fig. 2 present the partial scatterplots between the task-induced change in RSA and SE and SI, respectively. The scatterplot between the SE and the task-induced change in RSA. Fig. 1. The scatterplot between the SE and the task-induced change in RSA. Figure options The scatterplot between the SI and the task-induced change in RSA. Fig. 2. The scatterplot between the SI and the task-induced change in RSA.