الگوهای واکنش به استرس کورتیزول آزمایشگاهی
|کد مقاله||سال انتشار||مقاله انگلیسی||ترجمه فارسی||تعداد کلمات|
|39034||2004||10 صفحه PDF||سفارش دهید||محاسبه نشده|
Publisher : Elsevier - Science Direct (الزویر - ساینس دایرکت)
Journal : Hormones and Behavior, Volume 46, Issue 5, December 2004, Pages 618–627
Abstract Cortisol responses to a laboratory stress protocol were investigated in 82 male firefighters. Saliva samples were collected during an adaptation period beginning between 9 and 10 am, and then at the end of each of six 10-min trials (a mental arithmetic task, an inter-task recovery period, a speech task, and three recovery periods). Individual differences in the mean cortisol response to the stress tasks were characterized by variation in the direction of the response, as well as the size of the response. Neither pre-stress cortisol levels nor responses were correlated with cardiovascular and mood responses. Cortisol levels before stress task presentation were negatively correlated with recent stress severity. Larger mean cortisol responses were associated with lower reports of recent stress exposure, lower negative affect scores, and a coping style characterized less experience of anger, more control over anger expression, and a tendency to screen out threatening information in stressful situations. Thus, increased cortisol activity was associated with less recent stress exposure and a more adaptive behavioral style than for those whose cortisol levels fell or were largely unchanged in response to a laboratory stressor.
Introduction Increased cortisol activity has been demonstrated in response to laboratory and daily life stress (e.g., Brantley et al., 1988, Deinzer et al., 1997 and Kirschbaum et al., 1992). Generally, the more salient and intense a stimulus, the larger the response, thus parachute jumping elicits far larger responses than public mental arithmetic or speech task performance (Deinzer et al., 1997 vs. Kirschbaum et al., 1992). However, there are some inconsistencies with a simple linear stress–response relationship. Cortisol decreases during stress in some individuals (Heim et al., 2000), and lower cortisol levels with higher recent stress exposure (Roy et al., 2003). Differences in the pattern of cortisol responses also introduce significant methodological problems (e.g., Deinzer et al., 1997), and there are inconsistencies concerning cortisol's association with other reactivity measures and measures of affect and coping. Cortisol regulation and stress Individual differences in part reflect the state of hypothalamic pituitary adrenal (HPA) regulation (Keller-Wood and Dallman, 1984), in turn modulated by antecedent events. However, cortisol is ubiquitous, and thus regulation serves diverse roles, for example, as an immune regulator (McEwen et al., 1997), as a promoter of energy mobilization (Sapolsky, 1992), and a potentiator of vascular catecholamine action (Walker and Williams (1992). Recent evidence has shown that cortisol responding is not unidirectional (Heim et al., 2000), and hypocortisolism describes a profile where there is a reduced mean output, hyporeactivity, enhanced negative feedback, or reduced receptor sensitivity. Munck et al. (1984) suggested that cortisol promotes equilibrium during stress (an allostatic agent), preventing an overshoot in the defense reaction and conserving resources. But one caveat is that serving one homeostatic goal may ultimately compromise another (e.g., potentiating catecholamine vasoconstriction), and thus generate allostatic load (McEwen and Seeman, 1999). Hypocortisolism as a component of the PTSD response profile (Yehuda et al., 1993), may be functionally analogous to the inadequate response allostatic load profile described by McEwen and Seeman (1999). Variation in the patterns of cortisol responses represents an interesting challenge for the analysis of individual differences. Firstly, change scores are insensitive to changes that would have occurred simply as a function of diurnal variation, a problem also for response algorithms such as area under the curve. As the diurnal pattern is absent or unstable in almost 50% of individuals (Smyth et al., 1997), sampling on a control day is not reliable. Variation in the timing of responses also presents a problem (Deinzer et al., 1997 and Richter et al., 1996), as does variation in the direction of the cortisol change (Heim et al., 2000 and Rose and Fogg, 1993). Lastly, Kirschbaum et al. (1995) found that reactivity across five repeated daily stress sessions did not habituate in only 35%. Generalization of responses across stress axes Some studies have found that high cardiovascular reactors exhibit larger adrenocortical responses (Cacioppo et al., 1995 and Sgoutas Emch et al., 1994). However, Lovallo et al. (1990) found this was limited to aversive stressors. al'Absi et al. (1994) found that borderline hypertensives showed larger cortisol responses than normotensives to continuous but not intermittent stress tasks. Thus, greater autonomic reactivity appeared to be associated with greater cortisol reactivity, but only for certain stress task characteristics. Affect, coping, and recent stress exposure, and their association with cortisol responses Various factors may contribute to inconsistent findings, including operationalization of cortisol activity (e.g., level, reactivity), moderating factors (e.g., coping styles or recent stress exposure history). Ellenbogen et al. (2001) found depression was differentially associated with stress response change scores (negatively correlated) and the recovery level (positively correlated). Roy et al. (2001) found cortisol recovery (recovery-stress) was not associated with depression or anxiety, and Young and Nolen-Hoeksema (2001) found no association between cortisol responses and tendency to ruminate (passive and repetitive focus on distress). Pruessner et al. (1997) raise a methodological issue, finding reliable associations only when cortisol responses were averaged over multiple sessions. Scarpa and Luscher (2002) found self-esteem differentially modulated the association with depression, a cortisol fall predicting low self-esteem depression, an increase predicting high self-esteem depression, and the highest depression scores predicted by a cortisol fall and low self-esteem. A number of studies have reported a negative association between measures of recent stress exposure and current cortisol activity, particularly where there is extreme variation in exposure to stress (e.g., PTSD, Yehuda et al., 1993; combat Bourne et al., 1967), but also in response to relatively common fluctuations in daily life stress exposure (e.g., Roy et al., 2003). Coping (disengagement) has also been implicated in the hypocortisolism response to stress exposure described in PTSD (Mason et al., 2001). The primary aim of this paper was to explore sources of individual difference in the laboratory stress salivary cortisol response. The presumption that cortisol levels increase under conditions of stress seems to be violated by results from several studies. It is hypothesized that laboratory stress will be associated with an increase in cortisol in some participants, little change in some, and a reduction in the cortisol level in others. Analyses explore whether cortisol responses were associated with cortisol levels before the stress challenge. The study investigated whether changes in cortisol activity were associated with changes in cardiovascular and mood responses. Due to the inconclusive nature of previous studies, it was hypothesized that there would be no generalized reactivity pattern across measures (cortisol, cardiovascular, and mood responses) and that change scores would be uncorrelated. Finally, the study investigated the association between cortisol activity and measures of affect, coping, and recent stress exposure.
نتیجه گیری انگلیسی
Results Background characteristics The average age was 25 years (19–31, SEM 0.4), and their average body mass index was 24.4 (19.1–29.3). All participants were normotensive; however, 35.6% of subjects reported a parental cardiovascular illness. Thirty-four reported being either occasional or frequent smokers, but a χ2 analysis revealed no significant association between smoking status (smoker vs. non-smoker) and cortisol response group (≥3 SEM fall, 1–3 SEM fall, ±1 SEM change, 1–3 SEM increase, and ≥3 SEM increase). Descriptive statistics for the questionnaires are shown in Table 1. Table 1. Affect, coping and recent stress inventory scores Mean SEM Range Measures of affect Mean BDI 3.7 0.4 0–20 Mean STAI-Y1 32 0.8 20–54 Anger expression Anger-In 16 0.4 10–24 Anger-Out 15 0.5 8–29 Anger control 22.6 0.6 10–32 Anger expression 24 .9 4–49 MBSS Monitoring 9 0.3 2–15 Blunting 3 0.2 0–9 Derogatis stress profile vectors Time pressure 13 0.4 6–22 Driven behavior 10 0.4 2–22 Attitude posture 15 0.4 7–25 Relaxation potential 12 0.5 4–21 Role definition 10 0.4 1–20 Vocational environment 9 0.4 1–21 Domestic environment 7 0.4 1–18 Health environment 8 0.4 1–17 Hostility 11 0.6 0–27 Depression 9.5 0.5 1–26 Anxiety 6 0.4 0–17 Total stress score 110 2.4 68–165 Job strain questionnaire Decision latitude 6 0.2 3–11 Job demand 8 0.2 4–12 Skill utilization 12 .2 9–16 Overall job strain 9 0.2 3.6–14 Recent stress reporting Hassle frequency 10 0.9 0–36 Hassle severity 1.4 0.05 0–2.3 DSI mean frequency 5 0.3 0.4–12 DSI mean severity 2 0.07 1–3.6 DSI peak severity 2.6 0.1 1–5 Table options Overall psychobiological stress responses A significant main effects of protocol trials was found for all physiological and mood response measures: SBP F(6, 510) = 126.45, P < 0.0001, γ = 1.0; DBP F(6, 510) = 149.76, P < 0.0001, γ = 1.0; HR F(6, 510) = 157.37, P < 0.0001, γ = 1.0; salivary cortisol, F(6, 522) = 9.03, P < 0.0001, γ = 1.0; AD ACL energetic activation, F(6, 528) = 99.77, P < 0.0001, γ = 1.0; and AD ACL tense activation, F(6, 528) = 113.58, P < 0.0001, γ = 1.0. Mean cardiovascular activity and mood responses were significantly higher during the stressor trials, fell toward baseline levels between stressor trials and had substantially recovered toward baseline levels within the first 10 min of the recovery period ( Fig. 1). An increase in cortisol activity was found at 10 min after each task had ended, that is, during the IT recovery period and during the R10 recovery period, though the mean cortisol response size was modest compared to some other studies employing quite extreme stressors (e.g., parachuting in Deinzer et al., 1997; and public stressor performance in Kirschbaum et al., 1992). As can be seen from the descriptive statistics for the mean cortisol change following the onset of the stressor tasks, (X¯ 0.18 nmol/l, SEM 0.13 nmol/l), the group as a whole showed a small increase, but considerable inter-individual variation, and the individual change scores ranged from a mean cortisol fall of 5.9 nmol/l to a mean cortisol increase of 2.5 nmol/l. In relative terms, this meant a mean percentage change of +11% (SEM 3.5%), ranging from a mean reduction of 55% to a mean increase of 124%. Neither the absolute mean cortisol change, nor the relative cortisol change was significantly correlated with either the cortisol level before the onset of the stress tasks, or the change in cortisol level between samples (pre-MA1 − pre-MA2). Overall cardiovascular, cortisol and mood responses to the laboratory stress ... Fig. 1. Overall cardiovascular, cortisol and mood responses to the laboratory stress protocol. Error bars represent standard error of the mean. Figure options Behavioral correlates pre-stress task cortisol activity The mean pre-mental arithmetic cortisol level (X¯ of pre-MA1 and pre-MA2) was negatively correlated with both the mean recent DSI severity rating (r = −0.245, P < 0.03) and the peak DSI severity rating (r = −0.244, P < 0.03), suggesting that higher reports of recent stress were associated with a lower cortisol level. There was no correlation with other pre-stress measures (baseline cardiovascular and tense mood measures), or dispositional measures (BDI; STAI-Y1; Anger Expression; and MBSS, DSP; Job Strain Questionnaire). Furthermore, the change in cortisol across the adaptation period (pre-MA1-pre-MA2) was uncorrelated with any behavioral, or cardiovascular and mood response measures. Behavioral correlates of the mean cortisol change score Mean cortisol change following the MA task was correlated with cardiovascular and tense mood change scores for the MA and ST trials (MA-BA and ST-BA), with dispositional measures (BDI; STAI-Y1; Anger Expression; and the MBSS), and measures of stress (DSP; Job Strain Questionnaire; DSI). Examining Table 2, several patterns seemed to be present in the correlations: (1) Dispositional measures (depression and anxiety measures) were negatively associated with the mean cortisol change, indicating that poorer affect was associated with a lower mean cortisol increase or a greater cortisol reduction. (2) Individuals who were more likely to experience anger (high Anger Expression), and or express that anger in a more aggressive way (high Anger-Out) were more likely to show smaller increases in cortisol activity and/or greater reductions in cortisol activity, while a propensity toward monitoring and preventing the experience and expression of anger (high Anger Control) was associated with larger cortisol increase. (3) Across a range of sub-scales from the Derogatis Stress Profile, and with convergent findings from other stress measures (DSI, Hassle Scale, and Job Strain Questionnaire), there was a consistent pattern of negative correlations with the mean change score. Thus, vulnerability to stress, greater experience of stress, and reporting of more negative consequences of stress were all associated with smaller increases in mean cortisol activity, or showed larger reductions in mean cortisol activity. Thus, those showing lower mean cortisol increases and/or greater reductions might endorse the achievement ethic (you get what you deserve), report more sources of stress from their domestic environment, report stronger affective experiences arising from anger, and have higher anxiety and depression scores. Furthermore, lower cortisol responses seemed to correspond to a higher reporting of recent stress exposure. Table 2. Correlates of the mean cortisol change following exposure to laboratory stress Negative correlates Positive correlates Affect measures (1) X¯ BDI (r = −0.35**) (2) X¯ State anxiety (r = −0.36**) Anger measures Anger measures (1) Anger/expression (r = −0.24*) (1) Anger/control (r = 0.22*) Derogatis stress profile (1) Attitude posture (r = −0.23*) (2) Domestic environment (r = −0.24*) (3) Hostility (r = −0.37**) (4) Anxiety (r = −0.29**) (5) Global stress score (r = −0.29**) Stress inventory (1) Hassles frequency (r = −0.27*) (2) Job demand (r = −0.24*) X¯ = The Beck Depression Inventory and Spielberger State Anxiety Questionnaires were completed five times over an 18-month period and the mean of these is used here. Significance level for the correlation coefficients is indicated by * = P < 0.05; ** = P < 0.01. Table options When the same correlations were run using the relative change in cortisol (mean percentage change) rather than the absolute response, the pattern of correlations was substantially the same. The size of the coefficients for correlations with depression (r = −0.22, P < 0.05) and anxiety (r = −0.22, P < 0.05) were slightly smaller, while the correlation with Anger Expression (r = −0.28, P < 0.01) and Anger Control (r = 0.33, P < 0.01) were slightly stronger, and there was an additional significant association with the blunting coping style (r = 0.25, P < 0.05). However, unlike mean cortisol change, the percentage cortisol change was not significantly correlated with either DSI, Hassles, or Job Strain measures of recent stress exposure, although the correlation with the Derogatis Global Stress Score was similar (r = −0.28, P < 0.05). Thus it appears that as far as the behavioral correlates examined in this paper, there is a substantial amount of overlap concerning associations with cortisol responses between a measure of the absolute cortisol response to a stressor and a measure of the relative cortisol response to a stressor.