- Email: [email protected]
The cortisol awakening response: Associations with trait anxiety and stress reactivity
Personality and Individual Differences 51 (2011) 123–127
Contents lists available at ScienceDirect
Personality and Individual Differences journal homepage: www.elsevier.com/locate/paid
The cortisol awakening response: Associations with trait anxiety and stress reactivity Sarah Walker a,⇑, Daryl B. O’Connor a,⇑, Alexandre Schaefer b, Duncan Talbot c, Hilde Hendrickx c a
Institute of Psychological Sciences, University of Leeds, Leeds LS2 9JT, UK Wolfson Research Institute, University of Durham, Stockton-on-Tees TS17 6BH, UK c Unilever Discover R&D, Colworth MK44 1LQ, UK b
a r t i c l e
i n f o
Article history: Received 12 October 2010 Received in revised form 3 March 2011 Accepted 16 March 2011 Available online 22 April 2011 Keywords: Stress Anxiety State Trait Salivary cortisol Cortisol awakening response
a b s t r a c t This study investigated the extent to which trait anxiety and state anxiety in response to stress are associated with the cortisol awakening response (CAR). Fifty-one healthy participants were recruited. State anxiety measures were taken in anticipation of and during a laboratory stressor. Salivary cortisol levels were measured immediately upon awakening (at 0, 15, 30, and 45 min) on two consecutive mornings. Cortisol awakening response was assessed by the area under the curve with respect to zero (AUCG). The magnitude of the CAR was found to be negatively associated with both trait anxiety and anticipatory anxiety. Moreover, regression analysis showed that the effects of trait anxiety on the AUCG were mediated by anticipatory anxiety. These results suggest that the CAR is influenced by trait anxiety. Moreover, the effect of trait anxiety on the CAR seems to operate by impacting on psychological stress reactivity (i.e., anticipatory anxiety). Ó 2011 Elsevier Ltd. All rights reserved.
1. Introduction Over the last 15 years the cortisol awakening response (CAR) has received a large amount of attention (O’Connor et al., 2009a). It is defined as the cortisol secretory response during the first 45–60 min immediately after awakening and recent research has indicated that the magnitude of the CAR is influenced by individual differences in personality. Previous research on the relationship between the CAR and personality has often focused on neuroticism related traits due to links with future health risk. Although subjective measures of health consistently suggest that neuroticism is associated with a range of self reported complaints and somatic symptoms (e.g., Taylor et al., 2008), the relationship between neuroticism and objective indices is inconsistent (Friedman & Booth-Kewley, 1987). These inconsistencies are evident within the CAR literature with findings indicating either that the CAR is blunted (e.g., Therrien et al., 2008), enhanced (e.g., Portella, Harmer, Flint, Cowen, & Goodwin, 2005), or even no different (e.g., van Santen et al., 2011) in individuals high in neuroticism (or other related trait factors) as compared to suitable comparison groups. To help elucidate these findings, it may be useful to explore potential mediating pathways. Research has postulated that ⇑ Corresponding authors. Tel.: +44 113 343 5727; fax: +44 113 343 5749. E-mail addresses: [email protected] (S. Walker), d.b.o’connor@leeds. ac.uk (D.B. O’Connor). 0191-8869/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.paid.2011.03.026
neuroticism can influence health by increasing reactivity towards stress (e.g., Bolger & Zuckerman, 1995) and that reactivity measures taken in anticipation of and during a stressor are able to differentiate those high and low in neuroticism related traits (e.g., Bolger & Zuckerman, 1995; Infrasca, 1997). As changes in cortisol have been found to occur in response to stress (see McEwen, 2007), it is possible that measures of psychological reactivity to stress may help explain the association between neuroticism and the CAR. For example, those high in neuroticism have previously been shown to exhibit a heightened psychological response to stress (Bolger & Zuckerman, 1995), which may have a knock on effect on hypothalamic pituitary adrenal (HPA) axis mediated stress responses. However, to the best of our knowledge, irrespective of the nature of the relationship, no studies have explored whether psychological stress reactivity is one of the mechanisms through which neuroticism related traits may influence the CAR. Moreover, it is important to note that neuroticism, as conceptualised within the five factor model of personality, is a broad construct which has been found to elicit mixed physiological results in terms of stress research (as discussed above). Therefore, in the current study the effects of trait anxiety were examined as this variable has been shown to be a more homogenous neuroticism-related dimension (see Luteijn & Bouman, 1988). When examining the impact of stressful encounters on the CAR, there has been large variability in the types of stressors explored, ranging from measures of day-to-day stressors/hassles (e.g., Adam,
124
S. Walker et al. / Personality and Individual Differences 51 (2011) 123–127
Hawkley, Kudielka, & Cacioppo, 2006), to psychological responses to laboratory-based stressors (e.g., Fabian et al., 2009). In the present study it was thought important to explore the impact of psychological responses to a laboratory-based stressful event on the CAR in order to standardise the stressful encounter that participants were exposed to. Therefore, here we explored the associations between psychological responses to negative images using stimuli from the International Affective Picture System (Lang, Bradley, & Cuthbert, 1999) and the CAR. In particular, two components of the psychological response to stress were measured (anticipatory anxiety: anxiety in anticipation of the stressor; task anxiety: anxiety during the stressor) as these components have previously been found to be significantly associated with cortisol responses to an acute stressor (e.g., Elzinga, Schmahl, Vermetten, van Dyck, & Bremner, 2003). Exploring the relationship between state anxiety and the CAR was used to provide an index of how someone’s general reaction to everyday stress may impact on cortisol output. A number of methodological and measurement issues may have accounted for the inconsistencies outlined in the previous personality, stress and CAR research findings. First, participant nonadherence to the sampling protocol is known to be a serious problem in CAR research and the importance of the timing of the samples has been found to be vital (O’Connor et al., 2009a; Thorn, Hucklebridge, Evans, & Clow, 2006). Second, it is essential to control for potential confounding variables. The effects of individual differences such as age, gender, awakening time, depression symptomatology and body mass index (BMI) on the CAR are inconsistent (for reviews see Almeida, Piazza, & Stawski, 2009; Fries, Dettenborn, & Kirschbaum, 2009). Moreover the directionality of any observed relationship is difficult to determine due to these mixed results. Nevertheless, the potential explanatory effect of such factors is thought important to consider in any CAR analysis. Therefore, in the current study, we removed participants suspected of non-adherence to the sampling protocol and examined the effects of age, gender, awakening time, BMI and depression symptomatology in all analyses. In summary, this study had two main aims: (1) to clarify the direction of the relationship between trait anxiety and the CAR, and (2) to explore whether the impact of trait anxiety is mediated through psychological stress reactivity (i.e., changes in state anxiety in anticipation and during a stressor).
2. Materials and methods 2.1. Design and participants Fifty-one healthy students and staff from a large university in England were recruited to take part in a study looking at the effects of emotional regulation on health. Participants were excluded if they suffered from any hormonal disorder, regularly used recreational drugs, were taking steroid-based or neurological/psychotropic medication, scored above 15 on the Beck Depression Inventory as this is the threshold for possible depression in community samples (BDI: Beck & Steer, 1987; Beck, Ward, Mendelson, Mock, & Erbaugh, 1961), or had been to see a psychologist/psychiatrist in the past 6 months. Forty participants (14 males and 26 females) were included in the final sample (see cortisol analysis section later) with a mean age of 24.4 years (7.06 SD) and BMI of 22.63 (3.32 SD). The sample size was informed by previous research investigating the relationship between personality and the CAR (Portella et al., 2005) and by a power calculation using G*Power 3 (Faul, Erdfelder, Lang, & Buchner, 2007). Effect size was set at d = .77, alpha at .05, power at .80. This recommended a total sample size of approximately 30 participants. Fifty-one were tested to allow for
outliers, drop outs and suspected non-adherence. Participants were reimbursed £20 for their time and inconvenience. A correlational design was utilised to analyse the association between trait anxiety and state anxiety in anticipation and during an acute laboratory stressor (day 1) and the CAR on two consecutive days (day 2 and day 3). 2.2. Procedures and measures All research was approved by the University Departmental Ethics Committee. Upon arrival on day 1 participants were asked to rest for 15 min before the stressor task commenced. To examine stress reactivity, the 6-item short form of the State-Trait Anxiety Inventory (STAI: Marteau & Bekker, 1992) was used. Participants were asked to rate how they feel right now (e.g., I feel calm) on a 4-point Likert scale ranging from 1 (not at all) to 4 (very much), with higher scores indicating higher levels of state anxiety. The STAI short form is a commonly used measure to assess state anxiety with research indicating that the reliability and validity of the scale is acceptable (Marteau & Bekker, 1992). This scale was administered immediately after the 15 min rest (‘‘anticipatory anxiety 1’’) and again approximately 20 min later during the stressor task set up (‘‘anticipatory anxiety 2’’), and 20 min into the stressor during a resting break which lasted approximately 1 min (‘‘task anxiety’’). In the present study, the scale for anticipatory anxiety 1, anticipatory anxiety 2 and task anxiety yielded good internal consistency with alphas of .72, .72 and .81, respectively. To examine the influence of trait anxiety and depression symptomatology, the trait version of the STAI (Spielberger, Gorsuch, Lushene, Vagg, & Jacobs, 1983) and the BDI (Beck et al., 1961) were also administered. The trait scale of the STAI is a 20-item measure used to assess trait levels of anxiety. Participants have to rate how they generally feel (e.g., I feel pleasant) on a 4-point Likert scale ranging from 1 (almost never) to 4 (almost always). The BDI is a 21-item scale used to analyse how one has been feeling in the past week. Each statement has at least four possible answer choices, which range in intensity. For both scales, higher scores indicate higher levels of trait anxiety or depression. Research indicates that the reliability and validity of the STAI-trait and BDI is acceptable (Beck & Steer, 1987; Rule & Traver, 1983). In the present study, the STAI and BDI yielded good internal consistency with alphas of .95 and .79, respectively. In order to assess the CAR, saliva samples were collected on two mornings (day 2 and day 3) following the laboratory stress task. The researcher guided participants in their selection of two suitable week days for the collection of the saliva samples. This was to prevent sampling on a weekend, as weekday versus weekend differences have been found to impact on the CAR (Clow, Thorn, Evans, & Hucklebridge, 2004). 2.2.1. Laboratory stress task The stressor task involved subjects being exposed to 120 negative images from the International Affective Picture System (IAPS: Lang et al., 1999) for 1 h. In terms of the negative images, only those which have previously been found to elicit high levels of negative affect and arousal were used (Lang et al., 1999). The negative pictures included images of dead and mutilated bodies, war scenes, and pictures of medical conditions. Participants were warned during the stressor task set up about the nature of these picture contents. Exposure to IAPS negative images is known to elicit anxiety (Dichter, Tomarken, & Baucom, 2002; Simmons, Matthews, Stein, & Paulus, 2004) which we confirmed in a pilot study where anxiety levels increased significantly after exposure to unpleasant IAPS images compared to before, t(25) = .3.20, p < .01. Therefore, two measures of each participant’s stress reactivity were computed and utilised in the current study: anticipatory anxiety (anticipatory anxiety 1 + 2) and task anxiety.
125
S. Walker et al. / Personality and Individual Differences 51 (2011) 123–127
2.2.2. Cortisol measurements Participants were asked to take their own saliva samples immediately upon awakening (whilst still in bed), at 15, 30, and 45 min thereafter on two mornings following the stressor task. Saliva was collected using polyester Salivette collection devices (Sarstedt, Germany). In order to improve compliance to the sampling protocol, participants were asked to report what time they woke up and what time they took each of the samples. They were also told the importance of taking each sample at the correct time and informed that the experimenter would be able to identify non-compliance by inspecting the CAR profiles. All subjects were asked to refrain from eating, drinking caffeine, alcohol, acidic drinks, smoking and brushing their teeth before and during sample taking (Newman, O’Connor, & Conner, 2007; Wust et al., 2000). Saliva samples were posted back and frozen within 7 days of sampling; which has previously been found to be an adequate amount of time for cortisol to remain at room temperature (Clements & Parker, 1998). 2.3. Cortisol analysis and suspected non-adherence All Salivettes were frozen at 80 °C until being sent to Unilever, Colworth UK for analysis. Salivette samples were centrifuged for 10 min, 400 g at 20 °C. Once decanted, the saliva samples were tested using a time-resolved competitive fluorescence immunoassay on an autoDELFIA (Perkin Elmer). All raw data were analysed and screened for suspected nonadherence following criteria described by Thorn et al. (2006). These authors have argued that it is possible to detect non-compliant participants by identifying individuals who show no rise in cortisol following awakening. This rationale is based on work by Kupper et al. (2005) who used recording devices to demonstrate that those who failed to exhibit a rise in cortisol woke up on average 42 min earlier than they reported. On this basis, five people were excluded due to a lack of rise in cortisol from awakening to either 15 or 30 min. Six additional participants also failed to complete the trait anxiety measure, leaving a total sample of 40 subjects for the statistical analysis. Self reported sample timings were also analysed to corroborate any suspected non-adherence. The data were averaged across the two days to gain a mean CAR for 67.5% of the sample, for the other 32.5% only one day of CAR measurement was analysed due to suspected non-adherence. Cortisol levels and trait anxiety were log transformed due to skewness.
Fig. 1. Mean (±SEM) awakening cortisol profile in participants low (lower tertile) and high (higher tertile) in trait anxiety.
at 30 min post awakening is observed, with a lower CAR apparent in those high in trait anxiety. Furthermore, the means shown in Table 1 indicate that anticipatory and task anxiety are of a similar level to those reported in other studies incorporating laboratory stressors (Willis, O’Connor, & Smith, 2005). A correlational analysis was conducted to explore the relationships between state anxiety in response to the stressor and trait anxiety on the AUCG (see Table 1). Trait anxiety and task anxiety were positively associated with anticipatory anxiety. However, only trait anxiety and anticipatory anxiety were negatively associated with the mean AUCG, indicating that lower levels of cortisol were associated with higher levels of trait anxiety and anticipatory anxiety. Therefore, to examine these effects further, we conducted a hierarchical regression analysis including only trait anxiety and anticipatory anxiety. 3.1. Predictors of the AUCG As shown in Table 2, trait anxiety was entered at Step 1 and found to account for 11% of the variance (b = .33, p < .05). Moreover, when entered in at Step 2, anticipatory anxiety accounted for a further 10% of the variance (b = .35, p < .05) and reduced the predictive value of trait anxiety to non significance suggesting mediation. Formal mediation analyses are reported below. 3.2. Mediation analysis
2.4. Statistical analysis Using the four sampling time points, the area under the curve with respect to zero (AUCG) was determined following established procedures (Pruessner, Kirschbaum, Meinlschmid, & Hellhammer, 2003).1 Pearson’s Product Moment correlation coefficients and hierarchical regression analysis were conducted using SPSS 16 to investigate the impact of trait anxiety on the AUCG. The effects of age, gender, awakening time, BMI and BDI score were examined and found to have no impact on the CAR; therefore these factors were not controlled for within the regression analysis. Trait anxiety was added at Step 1 and state anxiety in response to the stressor was added at Step 2. In all analyses, the mean AUCG was used as the main outcome variable.
For mediation to be demonstrated the following conditions need to be met (Kenny, Kashy, & Bolger, 1998): (1) the independent variable (i.e., trait anxiety) affects the mediator (i.e., anticipatory anxiety); (2) the independent variable affects the dependent variable (i.e., mean AUCG); (3) the mediator affects the dependent variable when the independent variable is controlled for and; (4) full mediation is confirmed when the association between the independent and dependent variable is reduced to non-significance after the effect of the mediator is controlled for. Therefore, an additional regression analysis was conducted and showed that trait
Table 1 Zero-order correlation coefficients, means and standard deviations for all main study variables.
3. Results
Mean 1. 2. 3. 4.
Figure 1 shows the CAR in those high (upper tertile) and low (lower tertile) in trait anxiety. A clear increase in cortisol peaking 1
The data were also analysed with regards to the area under the curve with respect to increase (AUCi), however no significant effects were found. Therefore, for the sake of brevity only the AUCG will be reported.
Trait anxiety Anticipatory anxiety Task anxiety AUCG
N = 40. p < .05. p < .01.
*
**
38.68 9.38 11.82 69.92
SD 11.75 2.02 3.29 16.69
1 _
2
3 **
.42 _
4 .35* .40* .31
.23 .39* _ _
126
S. Walker et al. / Personality and Individual Differences 51 (2011) 123–127
Table 2 Hierarchical multiple regression analyses testing the influence of trait and state anxiety on the CAR. Predictor Step 1: Trait anxiety Step 2: Anticipatory anxiety
b (step 1) *
.33
b (step 2) .17 .35*
DR2
Total R2
*
.11 .09*
.20
N = 40. * p < .05.
anxiety was predictive of anticipatory anxiety (b = .47, p < .01), accounting for 22% of the variance; confirming condition 1. As all other conditions were previously met (as above), together these results indicate that the effects of trait anxiety on the AUCG were mediated by anticipatory anxiety. To verify the mediation effects a Sobel test (Preacher & Leonardelli, 2001) was also conducted and mediation was confirmed (Z = 1.75, p < .05). 4. Discussion Two key findings have emerged from the current study. First, trait anxiety was found to be negatively associated with the CAR, indicating that individuals high in trait anxiety secrete less cortisol following awakening compared to individuals low in this trait. This result is contrary to some previous literature in the area (e.g., Schlotz, Hellhammer, Schulz, & Stone, 2004), however, it is broadly consistent with findings by Therrien et al. (2008), where the CAR has been found to be reduced in high trait anxious individuals. This also corroborates previous work which has found high trait anxiousness to be associated with an attenuated HPA axis response (Jezova, Makatsori, Duncko, Moncek, & Jakubek, 2004). Taken together, these findings point to a potential mechanism linking trait anxiety to dysregulation of the HPA axis. Individuals high in trait anxiety have been found to encounter a greater number of stressors and to react more negatively to stress (e.g., Bolger & Zuckerman, 1995). Therefore, over time, this excessive activation of the stress response may cause the HPA axis to become less responsive (cf., McEwen, 2007). This notion would be consistent with other recent findings whereby chronic stress has been found to be associated with decreased cortisol output (e.g., O’Connor et al., 2009a). Moreover, it has been suggested that alterations in HPA axis function may be due to improved sensitivity to the negative feedback of glucocorticoids or a reduction in key hormones from the hypothalamus leading to a reduction in cortisol release (O’Connor et al., 2009a), explaining the negative association found in the current study. Whether such mechanisms thought to regulate diurnal HPA axis function also regulate the CAR is unknown. However, it is likely that in cases where stress becomes chronic, this typically adaptive mechanism, over time, may no longer effectively modulate anticipated daily demands (Adam et al., 2006). Although speculative at present, it is possible that such changes in cortisol output may partly explain why individuals with high levels of trait anxiety, who may be more vulnerable to and experience a higher number of stressors, are at a greater risk of ill health (see Sherwood & Turner, 1992). Future research ought to test this possibility. Second, this study also found that one of the mechanisms by which trait anxiety impacts on the CAR may be through psychological stress reactivity (i.e., state anxiety responses). This mediating relationship is in accordance with previous work where it was found that in those already suffering from chronic stress/anxiety, reports of acute anxiety can exacerbate the effect of chronic anxiety on cortisol output (Miller & Chen, 2006). The effect of subjective indicators of emotion on the CAR has also been found elsewhere (e.g., Fabian et al., 2009), further reinforcing the importance of considering the impact of psychological responses
particularly when looking at the relationship between a trait factor and the CAR. It is acknowledged however, that this effect was found for anticipatory anxiety but not task anxiety. Although unexpected, these findings are in accordance with previous literature where anticipatory stress responses were found to be an important predictor of physiological outcomes (e.g., de Timary, Roy, Luminet, Fillée, & Mikolajczak, 2008). This implies that the anxiety felt towards an upcoming stressor is more likely to differentiate between those low and high in trait anxiety on physiological outcomes than the anxiety felt during the stressor itself. To date, this is the first study to find evidence for the potential mediating effects of psychological stress reactivity on the trait anxiety and CAR relationship. Future research should replicate the current findings and incorporate multiple measures of state anxiety in response to stressors. There are some limitations to the current study that require further comment. We acknowledge the importance of the issue of participant non-adherence to the sampling protocol. As a result, suspected non-adherence (SNA) analysis (Thorn et al., 2006) was utilised to reduce the likelihood that such a factor had a significant influence on the current findings. Nevertheless, future research may benefit from using electronic monitoring, which has been shown to improve sampling accuracy and to incorporate daily diary designs (Kudielka & Kirschbaum, 2003; O’Connor, Conner, Jones, McMillan, & Ferguson, 2009b; O’Connor, Jones, Conner, McMillan, & Ferguson, 2008). We also acknowledge that neuroticism is a broader dimension than trait anxiety and that these closely related traits have frequently been seen as functionally equivalent (e.g., Eysenck, 2004). It is therefore important that future research explores the extent to which individuals high in Big 5 measures of neuroticism exhibit responses similar to the high trait anxious participants in the current study. In conclusion, these results suggest that the CAR is influenced by trait anxiety, such that individuals high in trait anxiety secrete less cortisol following awakening compared to individuals low in trait anxiety. Moreover, the effect of trait anxiety on the CAR seems to operate by impacting on psychological stress reactivity.
References Adam, E. K., Hawkley, L. C., Kudielka, B. M., & Cacioppo, J. T. (2006). Day-to-day dynamics of experience-cortisol associations in a population-based sample of older adults. Proceedings of the National Academy of Sciences USA, 103, 17058–17063. Almeida, D. M., Piazza, J. R., & Stawski, R. S. (2009). Interindividual differences and intraindividual variability in the cortisol awakening response: An examination of age and gender. Psychology and Aging, 24, 819–827. Beck, A. T., & Steer, R. A. (1987). Beck depression inventory manual. San Antonio, TX: The Psychological Corporation. Beck, A. T., Ward, C. H., Mendelson, M., Mock, J., & Erbaugh, J. (1961). An inventory for measuring depression. Archives of General Psychiatry, 4, 561–571. Bolger, N., & Zuckerman, A. (1995). A framework for studying personality in the stress process. Journal of Personality and Social Psychology, 69, 890–902. Clements, A. D., & Parker, R. C. (1998). The relationship between salivary cortisol concentrations in frozen versus mailed samples. Psychoneuroendocrinology, 23, 613–616. Clow, A., Thorn, L., Evans, P., & Hucklebridge, F. (2004). The awakening cortisol response: Methodological issues and significance. Stress: The International Journal on the Biology of Stress, 7, 29–37. de Timary, P., Roy, E., Luminet, O., Fillée, C., & Mikolajczak, M. (2008). Relationship between alexithymia, alexithymia factors and salivary cortisol in men exposed to a social stress test. Psychoneuroendocrinology, 33, 1160–1164. Dichter, G. S., Tomarken, A. J., & Baucom, B. R. (2002). Startle modulation before, during and after exposure to emotional stimuli. International Journal of Psychophysiology, 43, 191–196. Elzinga, B. M., Schmahl, C. G., Vermetten, E., van Dyck, R., & Bremner, J. (2003). Higher cortisol levels following exposure to traumatic reminders in abuserelated PTSD. Neuropsychopharmacology, 28, 1656–1665. Eysenck, M. W. (2004). Trait anxiety, repressors and cognitive biases. In J. Yiend (Ed.), Cognition, emotion and psychopathology: Theoretical, empirical and clinical directions. Cambridge: Cambridge University Press. Fabian, L. A., McGuire, L., Page, G. G., Goodin, B. R., Edwards, R. R., & Haythornthwaite, J. (2009). The association of the cortisol awakening
S. Walker et al. / Personality and Individual Differences 51 (2011) 123–127 response with experimental pain ratings. Psychoneuroendocrinology, 34, 1247–1251. Faul, F., Erdfelder, E., Lang, A. G., & Buchner, A. (2007). G*Power 3: A flexible statistical power analysis for the social, behavioural, and biomedical sciences. Behaviour Research Methods, 39, 175–191. Friedman, H. S., & Booth-Kewley, S. (1987). The ‘disease-prone personality’: A metaanalytic view of the construct.. American Psychologist, 42, 539–555. Fries, E., Dettenborn, L., & Kirschbaum, C. (2009). The cortisol awakening response (CAR): Facts and future directions. International Journal of Psychophysiology, 72, 67–73. Infrasca, R. (1997). Alexithymia, neurovegetative arousal and neuroticism: An experimental study. Psychotherapy and Psychosomatics, 66, 276–280. Jezova, D., Makatsori, A., Duncko, R., Moncek, F., & Jakubek, M. (2004). High trait anxiety in healthy subjects is associated with low neuroendocrine activity during psychosocial stress. Progress Neuropsychopharmacology & Biological Psychiatry, 28, 1331–1336. Kenny, D. A., Kashy, D. A., & Bolger, N. (1998). Data analysis in social psychology. In D. T. Gilbert, S. T. Fiske, & G. Lindzey (Eds.), The handbook of social psychology. Boston, Massachusetts: McGraw-Hill Companies. Kudielka, B. M., & Kirschbaum, C. (2003). Awakening cortisol responses are influenced by health status and awakening time but not by menstrual cycle phase. Psychoneuroendocrinology, 28, 35–47. Kupper, N., de Geus, E. J., van den Berg, M., Kirschbaum, C., Boomsma, D. I., & Willemsen, G. (2005). Familial influences on basal salivary cortisol in an adult population. Psychoneuroendocrinology, 30, 857–868. Lang, P. J., Bradley, M. M., & Cuthbert, B. N. (1999). International affective picture system (IAPS): Instruction manual and affective ratings. Technical report A-4. The Center for Research in Psychophysiology, University of Florida, Gainesville, FL. Luteijn, F., & Bouman, T. K. (1988). The concepts of depression, anxiety, and neuroticism in questionnaires. European Journal of Personality, 2, 113–120. Marteau, T. M., & Bekker, H. (1992). The development of a six-item short-form of the state scale of the Spielberger state-trait anxiety inventory (STAI). British Journal of Clinical Psychology, 31, 301–306. McEwen, B. S. (2007). Physiology and neurobiology of stress and adaptation: Central role of the brain. Physiological Reviews, 87, 873–904. Miller, G. E., & Chen, E. (2006). Life stress and diminished expression of genes encoding the glucocorticoid receptor and b2-adrenergic receptor in children with asthma. Proceedings of the National Academy of Sciences of the United States of America, 103, 5496–5501. Newman, E., O’Connor, D. B., & Conner, M. T. (2007). Daily hassles and eating behaviour: The role of cortisol reactivity status. Psychoneuroendocrinology, 32, 125–132. O’Connor, D. B., Hendrickx, H., Dadd, T., Elliman, T., Willis, T. A., Talbot, D., et al. (2009a). Cortisol awakening rise in middle-aged women in relation to psychological stress. Psychoneuroendocrinology, 34, 1486–1494. O’Connor, D. B., Conner, M., Jones, F., McMillan, B., & Ferguson, E. (2009b). Exploring the benefits of conscientiousness: An investigation of daily stressors and health behaviors. Annals of Behavioral Medicine, 37, 184–196.
127
O’Connor, D. B., Jones, F., Conner, M., McMillan, B., & Ferguson, E. (2008). Effects of daily hassles and eating style on eating behaviour. Health Psychology, 27, S20–S31. Portella, M. J., Harmer, C. J., Flint, J., Cowen, P., & Goodwin, G. M. (2005). Enhanced early morning salivary cortisol in neuroticism. American Journal of Psychiatry, 162, 807–809. Preacher, K. L., & Leonardelli, G. J. (2001). Calculation for the Sobel test: An interactive calculation tool for mediation tests [Computer software]. Available from:. Pruessner, J. C., Kirschbaum, C., Meinlschmid, G., & Hellhammer, D. H. (2003). Two formulas for computation of the area under the curve represent measures of total hormone concentration versus time-dependent change. Psychoneuroendocrinology, 28, 916–931. Rule, W. R., & Traver, M. D. (1983). Test-retest reliabilities of state-trait anxiety inventory in a stressful social analogue situation. Journal of Personality Assessment, 47, 276–277. Schlotz, W., Hellhammer, J., Schulz, P., & Stone, A. A. (2004). Perceived work overload and chronic worrying predict weekend-weekday differences in the cortisol awakening response. Psychosomatic Medicine, 66, 207–214. Sherwood, A., & Turner, J. R. (1992). A conceptual and methodological overview of cardiovascular reactivity research. In A. S. J. R. Turner & K. C. Light (Eds.), Individual differences in cardiovascular response to stress (pp. 3–32). New York: Plenum. Simmons, A., Matthews, S. C., Stein, M. B., & Paulus, M. P. (2004). Anticipation of emotionally aversive visual stimuli activates right insula. Neuroreport, 15, 2261–2265. Spielberger, C. D., Gorsuch, R. L., Lushene, P. R., Vagg, P. R., & Jacobs, A. G. (Eds.). (1983). Manual for the state-trait anxiety inventory (form Y). Palo Alto: Consulting Psychologists Press, Inc. Taylor, M. K., Reis, J. P., Sausen, K. P., Padilla, G. A., Markham, A. E., Potterat, E. G., et al. (2008). Trait anxiety and salivary cortisol during free living and military stress. Aviation Space & Environmental Medicine, 79, 129–135. Therrien, F., Drapeau, V., Lupien, S. J., Beaulieu, S., Doré, J., Tremblay, A., et al. (2008). Awakening cortisol response in relation to psychosocial profiles and eating behaviors. Physiology & Behavior, 93, 282–288. Thorn, L., Hucklebridge, F., Evans, P., & Clow, A. (2006). Suspected non-adherence and weekend versus week day differences in the awakening cortisol response. Psychoneuroendocrinology, 31, 1009–1018. van Santen, A., Vreeburg, S. A., Van der Does, A. J. W., Spinhoven, P., Zitman, F. G., & Penninx, B. W. J. H. (2011). Psychological traits and the cortisol awakening response: Results from the Netherlands study of depression and anxiety. Psychoneuroendocrinology, 36, 240–248. Willis, T. A., O’Connor, D. B., & Smith, L. (2005). The influence of morningness– eveningness on anxiety and cardiovascular responses to stress. Physiology & Behavior, 85, 125–133. Wust, S., Wolf, J., Hellhammer, D. H., Federenko, I., Schommer, N., & Kirschbaum, C. (2000). The cortisol awakening response – Normal values and confounds. Noise Health, 2, 79–88.
Contents lists available at ScienceDirect
Personality and Individual Differences journal homepage: www.elsevier.com/locate/paid
The cortisol awakening response: Associations with trait anxiety and stress reactivity Sarah Walker a,⇑, Daryl B. O’Connor a,⇑, Alexandre Schaefer b, Duncan Talbot c, Hilde Hendrickx c a
Institute of Psychological Sciences, University of Leeds, Leeds LS2 9JT, UK Wolfson Research Institute, University of Durham, Stockton-on-Tees TS17 6BH, UK c Unilever Discover R&D, Colworth MK44 1LQ, UK b
a r t i c l e
i n f o
Article history: Received 12 October 2010 Received in revised form 3 March 2011 Accepted 16 March 2011 Available online 22 April 2011 Keywords: Stress Anxiety State Trait Salivary cortisol Cortisol awakening response
a b s t r a c t This study investigated the extent to which trait anxiety and state anxiety in response to stress are associated with the cortisol awakening response (CAR). Fifty-one healthy participants were recruited. State anxiety measures were taken in anticipation of and during a laboratory stressor. Salivary cortisol levels were measured immediately upon awakening (at 0, 15, 30, and 45 min) on two consecutive mornings. Cortisol awakening response was assessed by the area under the curve with respect to zero (AUCG). The magnitude of the CAR was found to be negatively associated with both trait anxiety and anticipatory anxiety. Moreover, regression analysis showed that the effects of trait anxiety on the AUCG were mediated by anticipatory anxiety. These results suggest that the CAR is influenced by trait anxiety. Moreover, the effect of trait anxiety on the CAR seems to operate by impacting on psychological stress reactivity (i.e., anticipatory anxiety). Ó 2011 Elsevier Ltd. All rights reserved.
1. Introduction Over the last 15 years the cortisol awakening response (CAR) has received a large amount of attention (O’Connor et al., 2009a). It is defined as the cortisol secretory response during the first 45–60 min immediately after awakening and recent research has indicated that the magnitude of the CAR is influenced by individual differences in personality. Previous research on the relationship between the CAR and personality has often focused on neuroticism related traits due to links with future health risk. Although subjective measures of health consistently suggest that neuroticism is associated with a range of self reported complaints and somatic symptoms (e.g., Taylor et al., 2008), the relationship between neuroticism and objective indices is inconsistent (Friedman & Booth-Kewley, 1987). These inconsistencies are evident within the CAR literature with findings indicating either that the CAR is blunted (e.g., Therrien et al., 2008), enhanced (e.g., Portella, Harmer, Flint, Cowen, & Goodwin, 2005), or even no different (e.g., van Santen et al., 2011) in individuals high in neuroticism (or other related trait factors) as compared to suitable comparison groups. To help elucidate these findings, it may be useful to explore potential mediating pathways. Research has postulated that ⇑ Corresponding authors. Tel.: +44 113 343 5727; fax: +44 113 343 5749. E-mail addresses: [email protected] (S. Walker), d.b.o’connor@leeds. ac.uk (D.B. O’Connor). 0191-8869/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.paid.2011.03.026
neuroticism can influence health by increasing reactivity towards stress (e.g., Bolger & Zuckerman, 1995) and that reactivity measures taken in anticipation of and during a stressor are able to differentiate those high and low in neuroticism related traits (e.g., Bolger & Zuckerman, 1995; Infrasca, 1997). As changes in cortisol have been found to occur in response to stress (see McEwen, 2007), it is possible that measures of psychological reactivity to stress may help explain the association between neuroticism and the CAR. For example, those high in neuroticism have previously been shown to exhibit a heightened psychological response to stress (Bolger & Zuckerman, 1995), which may have a knock on effect on hypothalamic pituitary adrenal (HPA) axis mediated stress responses. However, to the best of our knowledge, irrespective of the nature of the relationship, no studies have explored whether psychological stress reactivity is one of the mechanisms through which neuroticism related traits may influence the CAR. Moreover, it is important to note that neuroticism, as conceptualised within the five factor model of personality, is a broad construct which has been found to elicit mixed physiological results in terms of stress research (as discussed above). Therefore, in the current study the effects of trait anxiety were examined as this variable has been shown to be a more homogenous neuroticism-related dimension (see Luteijn & Bouman, 1988). When examining the impact of stressful encounters on the CAR, there has been large variability in the types of stressors explored, ranging from measures of day-to-day stressors/hassles (e.g., Adam,
124
S. Walker et al. / Personality and Individual Differences 51 (2011) 123–127
Hawkley, Kudielka, & Cacioppo, 2006), to psychological responses to laboratory-based stressors (e.g., Fabian et al., 2009). In the present study it was thought important to explore the impact of psychological responses to a laboratory-based stressful event on the CAR in order to standardise the stressful encounter that participants were exposed to. Therefore, here we explored the associations between psychological responses to negative images using stimuli from the International Affective Picture System (Lang, Bradley, & Cuthbert, 1999) and the CAR. In particular, two components of the psychological response to stress were measured (anticipatory anxiety: anxiety in anticipation of the stressor; task anxiety: anxiety during the stressor) as these components have previously been found to be significantly associated with cortisol responses to an acute stressor (e.g., Elzinga, Schmahl, Vermetten, van Dyck, & Bremner, 2003). Exploring the relationship between state anxiety and the CAR was used to provide an index of how someone’s general reaction to everyday stress may impact on cortisol output. A number of methodological and measurement issues may have accounted for the inconsistencies outlined in the previous personality, stress and CAR research findings. First, participant nonadherence to the sampling protocol is known to be a serious problem in CAR research and the importance of the timing of the samples has been found to be vital (O’Connor et al., 2009a; Thorn, Hucklebridge, Evans, & Clow, 2006). Second, it is essential to control for potential confounding variables. The effects of individual differences such as age, gender, awakening time, depression symptomatology and body mass index (BMI) on the CAR are inconsistent (for reviews see Almeida, Piazza, & Stawski, 2009; Fries, Dettenborn, & Kirschbaum, 2009). Moreover the directionality of any observed relationship is difficult to determine due to these mixed results. Nevertheless, the potential explanatory effect of such factors is thought important to consider in any CAR analysis. Therefore, in the current study, we removed participants suspected of non-adherence to the sampling protocol and examined the effects of age, gender, awakening time, BMI and depression symptomatology in all analyses. In summary, this study had two main aims: (1) to clarify the direction of the relationship between trait anxiety and the CAR, and (2) to explore whether the impact of trait anxiety is mediated through psychological stress reactivity (i.e., changes in state anxiety in anticipation and during a stressor).
2. Materials and methods 2.1. Design and participants Fifty-one healthy students and staff from a large university in England were recruited to take part in a study looking at the effects of emotional regulation on health. Participants were excluded if they suffered from any hormonal disorder, regularly used recreational drugs, were taking steroid-based or neurological/psychotropic medication, scored above 15 on the Beck Depression Inventory as this is the threshold for possible depression in community samples (BDI: Beck & Steer, 1987; Beck, Ward, Mendelson, Mock, & Erbaugh, 1961), or had been to see a psychologist/psychiatrist in the past 6 months. Forty participants (14 males and 26 females) were included in the final sample (see cortisol analysis section later) with a mean age of 24.4 years (7.06 SD) and BMI of 22.63 (3.32 SD). The sample size was informed by previous research investigating the relationship between personality and the CAR (Portella et al., 2005) and by a power calculation using G*Power 3 (Faul, Erdfelder, Lang, & Buchner, 2007). Effect size was set at d = .77, alpha at .05, power at .80. This recommended a total sample size of approximately 30 participants. Fifty-one were tested to allow for
outliers, drop outs and suspected non-adherence. Participants were reimbursed £20 for their time and inconvenience. A correlational design was utilised to analyse the association between trait anxiety and state anxiety in anticipation and during an acute laboratory stressor (day 1) and the CAR on two consecutive days (day 2 and day 3). 2.2. Procedures and measures All research was approved by the University Departmental Ethics Committee. Upon arrival on day 1 participants were asked to rest for 15 min before the stressor task commenced. To examine stress reactivity, the 6-item short form of the State-Trait Anxiety Inventory (STAI: Marteau & Bekker, 1992) was used. Participants were asked to rate how they feel right now (e.g., I feel calm) on a 4-point Likert scale ranging from 1 (not at all) to 4 (very much), with higher scores indicating higher levels of state anxiety. The STAI short form is a commonly used measure to assess state anxiety with research indicating that the reliability and validity of the scale is acceptable (Marteau & Bekker, 1992). This scale was administered immediately after the 15 min rest (‘‘anticipatory anxiety 1’’) and again approximately 20 min later during the stressor task set up (‘‘anticipatory anxiety 2’’), and 20 min into the stressor during a resting break which lasted approximately 1 min (‘‘task anxiety’’). In the present study, the scale for anticipatory anxiety 1, anticipatory anxiety 2 and task anxiety yielded good internal consistency with alphas of .72, .72 and .81, respectively. To examine the influence of trait anxiety and depression symptomatology, the trait version of the STAI (Spielberger, Gorsuch, Lushene, Vagg, & Jacobs, 1983) and the BDI (Beck et al., 1961) were also administered. The trait scale of the STAI is a 20-item measure used to assess trait levels of anxiety. Participants have to rate how they generally feel (e.g., I feel pleasant) on a 4-point Likert scale ranging from 1 (almost never) to 4 (almost always). The BDI is a 21-item scale used to analyse how one has been feeling in the past week. Each statement has at least four possible answer choices, which range in intensity. For both scales, higher scores indicate higher levels of trait anxiety or depression. Research indicates that the reliability and validity of the STAI-trait and BDI is acceptable (Beck & Steer, 1987; Rule & Traver, 1983). In the present study, the STAI and BDI yielded good internal consistency with alphas of .95 and .79, respectively. In order to assess the CAR, saliva samples were collected on two mornings (day 2 and day 3) following the laboratory stress task. The researcher guided participants in their selection of two suitable week days for the collection of the saliva samples. This was to prevent sampling on a weekend, as weekday versus weekend differences have been found to impact on the CAR (Clow, Thorn, Evans, & Hucklebridge, 2004). 2.2.1. Laboratory stress task The stressor task involved subjects being exposed to 120 negative images from the International Affective Picture System (IAPS: Lang et al., 1999) for 1 h. In terms of the negative images, only those which have previously been found to elicit high levels of negative affect and arousal were used (Lang et al., 1999). The negative pictures included images of dead and mutilated bodies, war scenes, and pictures of medical conditions. Participants were warned during the stressor task set up about the nature of these picture contents. Exposure to IAPS negative images is known to elicit anxiety (Dichter, Tomarken, & Baucom, 2002; Simmons, Matthews, Stein, & Paulus, 2004) which we confirmed in a pilot study where anxiety levels increased significantly after exposure to unpleasant IAPS images compared to before, t(25) = .3.20, p < .01. Therefore, two measures of each participant’s stress reactivity were computed and utilised in the current study: anticipatory anxiety (anticipatory anxiety 1 + 2) and task anxiety.
125
S. Walker et al. / Personality and Individual Differences 51 (2011) 123–127
2.2.2. Cortisol measurements Participants were asked to take their own saliva samples immediately upon awakening (whilst still in bed), at 15, 30, and 45 min thereafter on two mornings following the stressor task. Saliva was collected using polyester Salivette collection devices (Sarstedt, Germany). In order to improve compliance to the sampling protocol, participants were asked to report what time they woke up and what time they took each of the samples. They were also told the importance of taking each sample at the correct time and informed that the experimenter would be able to identify non-compliance by inspecting the CAR profiles. All subjects were asked to refrain from eating, drinking caffeine, alcohol, acidic drinks, smoking and brushing their teeth before and during sample taking (Newman, O’Connor, & Conner, 2007; Wust et al., 2000). Saliva samples were posted back and frozen within 7 days of sampling; which has previously been found to be an adequate amount of time for cortisol to remain at room temperature (Clements & Parker, 1998). 2.3. Cortisol analysis and suspected non-adherence All Salivettes were frozen at 80 °C until being sent to Unilever, Colworth UK for analysis. Salivette samples were centrifuged for 10 min, 400 g at 20 °C. Once decanted, the saliva samples were tested using a time-resolved competitive fluorescence immunoassay on an autoDELFIA (Perkin Elmer). All raw data were analysed and screened for suspected nonadherence following criteria described by Thorn et al. (2006). These authors have argued that it is possible to detect non-compliant participants by identifying individuals who show no rise in cortisol following awakening. This rationale is based on work by Kupper et al. (2005) who used recording devices to demonstrate that those who failed to exhibit a rise in cortisol woke up on average 42 min earlier than they reported. On this basis, five people were excluded due to a lack of rise in cortisol from awakening to either 15 or 30 min. Six additional participants also failed to complete the trait anxiety measure, leaving a total sample of 40 subjects for the statistical analysis. Self reported sample timings were also analysed to corroborate any suspected non-adherence. The data were averaged across the two days to gain a mean CAR for 67.5% of the sample, for the other 32.5% only one day of CAR measurement was analysed due to suspected non-adherence. Cortisol levels and trait anxiety were log transformed due to skewness.
Fig. 1. Mean (±SEM) awakening cortisol profile in participants low (lower tertile) and high (higher tertile) in trait anxiety.
at 30 min post awakening is observed, with a lower CAR apparent in those high in trait anxiety. Furthermore, the means shown in Table 1 indicate that anticipatory and task anxiety are of a similar level to those reported in other studies incorporating laboratory stressors (Willis, O’Connor, & Smith, 2005). A correlational analysis was conducted to explore the relationships between state anxiety in response to the stressor and trait anxiety on the AUCG (see Table 1). Trait anxiety and task anxiety were positively associated with anticipatory anxiety. However, only trait anxiety and anticipatory anxiety were negatively associated with the mean AUCG, indicating that lower levels of cortisol were associated with higher levels of trait anxiety and anticipatory anxiety. Therefore, to examine these effects further, we conducted a hierarchical regression analysis including only trait anxiety and anticipatory anxiety. 3.1. Predictors of the AUCG As shown in Table 2, trait anxiety was entered at Step 1 and found to account for 11% of the variance (b = .33, p < .05). Moreover, when entered in at Step 2, anticipatory anxiety accounted for a further 10% of the variance (b = .35, p < .05) and reduced the predictive value of trait anxiety to non significance suggesting mediation. Formal mediation analyses are reported below. 3.2. Mediation analysis
2.4. Statistical analysis Using the four sampling time points, the area under the curve with respect to zero (AUCG) was determined following established procedures (Pruessner, Kirschbaum, Meinlschmid, & Hellhammer, 2003).1 Pearson’s Product Moment correlation coefficients and hierarchical regression analysis were conducted using SPSS 16 to investigate the impact of trait anxiety on the AUCG. The effects of age, gender, awakening time, BMI and BDI score were examined and found to have no impact on the CAR; therefore these factors were not controlled for within the regression analysis. Trait anxiety was added at Step 1 and state anxiety in response to the stressor was added at Step 2. In all analyses, the mean AUCG was used as the main outcome variable.
For mediation to be demonstrated the following conditions need to be met (Kenny, Kashy, & Bolger, 1998): (1) the independent variable (i.e., trait anxiety) affects the mediator (i.e., anticipatory anxiety); (2) the independent variable affects the dependent variable (i.e., mean AUCG); (3) the mediator affects the dependent variable when the independent variable is controlled for and; (4) full mediation is confirmed when the association between the independent and dependent variable is reduced to non-significance after the effect of the mediator is controlled for. Therefore, an additional regression analysis was conducted and showed that trait
Table 1 Zero-order correlation coefficients, means and standard deviations for all main study variables.
3. Results
Mean 1. 2. 3. 4.
Figure 1 shows the CAR in those high (upper tertile) and low (lower tertile) in trait anxiety. A clear increase in cortisol peaking 1
The data were also analysed with regards to the area under the curve with respect to increase (AUCi), however no significant effects were found. Therefore, for the sake of brevity only the AUCG will be reported.
Trait anxiety Anticipatory anxiety Task anxiety AUCG
N = 40. p < .05. p < .01.
*
**
38.68 9.38 11.82 69.92
SD 11.75 2.02 3.29 16.69
1 _
2
3 **
.42 _
4 .35* .40* .31
.23 .39* _ _
126
S. Walker et al. / Personality and Individual Differences 51 (2011) 123–127
Table 2 Hierarchical multiple regression analyses testing the influence of trait and state anxiety on the CAR. Predictor Step 1: Trait anxiety Step 2: Anticipatory anxiety
b (step 1) *
.33
b (step 2) .17 .35*
DR2
Total R2
*
.11 .09*
.20
N = 40. * p < .05.
anxiety was predictive of anticipatory anxiety (b = .47, p < .01), accounting for 22% of the variance; confirming condition 1. As all other conditions were previously met (as above), together these results indicate that the effects of trait anxiety on the AUCG were mediated by anticipatory anxiety. To verify the mediation effects a Sobel test (Preacher & Leonardelli, 2001) was also conducted and mediation was confirmed (Z = 1.75, p < .05). 4. Discussion Two key findings have emerged from the current study. First, trait anxiety was found to be negatively associated with the CAR, indicating that individuals high in trait anxiety secrete less cortisol following awakening compared to individuals low in this trait. This result is contrary to some previous literature in the area (e.g., Schlotz, Hellhammer, Schulz, & Stone, 2004), however, it is broadly consistent with findings by Therrien et al. (2008), where the CAR has been found to be reduced in high trait anxious individuals. This also corroborates previous work which has found high trait anxiousness to be associated with an attenuated HPA axis response (Jezova, Makatsori, Duncko, Moncek, & Jakubek, 2004). Taken together, these findings point to a potential mechanism linking trait anxiety to dysregulation of the HPA axis. Individuals high in trait anxiety have been found to encounter a greater number of stressors and to react more negatively to stress (e.g., Bolger & Zuckerman, 1995). Therefore, over time, this excessive activation of the stress response may cause the HPA axis to become less responsive (cf., McEwen, 2007). This notion would be consistent with other recent findings whereby chronic stress has been found to be associated with decreased cortisol output (e.g., O’Connor et al., 2009a). Moreover, it has been suggested that alterations in HPA axis function may be due to improved sensitivity to the negative feedback of glucocorticoids or a reduction in key hormones from the hypothalamus leading to a reduction in cortisol release (O’Connor et al., 2009a), explaining the negative association found in the current study. Whether such mechanisms thought to regulate diurnal HPA axis function also regulate the CAR is unknown. However, it is likely that in cases where stress becomes chronic, this typically adaptive mechanism, over time, may no longer effectively modulate anticipated daily demands (Adam et al., 2006). Although speculative at present, it is possible that such changes in cortisol output may partly explain why individuals with high levels of trait anxiety, who may be more vulnerable to and experience a higher number of stressors, are at a greater risk of ill health (see Sherwood & Turner, 1992). Future research ought to test this possibility. Second, this study also found that one of the mechanisms by which trait anxiety impacts on the CAR may be through psychological stress reactivity (i.e., state anxiety responses). This mediating relationship is in accordance with previous work where it was found that in those already suffering from chronic stress/anxiety, reports of acute anxiety can exacerbate the effect of chronic anxiety on cortisol output (Miller & Chen, 2006). The effect of subjective indicators of emotion on the CAR has also been found elsewhere (e.g., Fabian et al., 2009), further reinforcing the importance of considering the impact of psychological responses
particularly when looking at the relationship between a trait factor and the CAR. It is acknowledged however, that this effect was found for anticipatory anxiety but not task anxiety. Although unexpected, these findings are in accordance with previous literature where anticipatory stress responses were found to be an important predictor of physiological outcomes (e.g., de Timary, Roy, Luminet, Fillée, & Mikolajczak, 2008). This implies that the anxiety felt towards an upcoming stressor is more likely to differentiate between those low and high in trait anxiety on physiological outcomes than the anxiety felt during the stressor itself. To date, this is the first study to find evidence for the potential mediating effects of psychological stress reactivity on the trait anxiety and CAR relationship. Future research should replicate the current findings and incorporate multiple measures of state anxiety in response to stressors. There are some limitations to the current study that require further comment. We acknowledge the importance of the issue of participant non-adherence to the sampling protocol. As a result, suspected non-adherence (SNA) analysis (Thorn et al., 2006) was utilised to reduce the likelihood that such a factor had a significant influence on the current findings. Nevertheless, future research may benefit from using electronic monitoring, which has been shown to improve sampling accuracy and to incorporate daily diary designs (Kudielka & Kirschbaum, 2003; O’Connor, Conner, Jones, McMillan, & Ferguson, 2009b; O’Connor, Jones, Conner, McMillan, & Ferguson, 2008). We also acknowledge that neuroticism is a broader dimension than trait anxiety and that these closely related traits have frequently been seen as functionally equivalent (e.g., Eysenck, 2004). It is therefore important that future research explores the extent to which individuals high in Big 5 measures of neuroticism exhibit responses similar to the high trait anxious participants in the current study. In conclusion, these results suggest that the CAR is influenced by trait anxiety, such that individuals high in trait anxiety secrete less cortisol following awakening compared to individuals low in trait anxiety. Moreover, the effect of trait anxiety on the CAR seems to operate by impacting on psychological stress reactivity.
References Adam, E. K., Hawkley, L. C., Kudielka, B. M., & Cacioppo, J. T. (2006). Day-to-day dynamics of experience-cortisol associations in a population-based sample of older adults. Proceedings of the National Academy of Sciences USA, 103, 17058–17063. Almeida, D. M., Piazza, J. R., & Stawski, R. S. (2009). Interindividual differences and intraindividual variability in the cortisol awakening response: An examination of age and gender. Psychology and Aging, 24, 819–827. Beck, A. T., & Steer, R. A. (1987). Beck depression inventory manual. San Antonio, TX: The Psychological Corporation. Beck, A. T., Ward, C. H., Mendelson, M., Mock, J., & Erbaugh, J. (1961). An inventory for measuring depression. Archives of General Psychiatry, 4, 561–571. Bolger, N., & Zuckerman, A. (1995). A framework for studying personality in the stress process. Journal of Personality and Social Psychology, 69, 890–902. Clements, A. D., & Parker, R. C. (1998). The relationship between salivary cortisol concentrations in frozen versus mailed samples. Psychoneuroendocrinology, 23, 613–616. Clow, A., Thorn, L., Evans, P., & Hucklebridge, F. (2004). The awakening cortisol response: Methodological issues and significance. Stress: The International Journal on the Biology of Stress, 7, 29–37. de Timary, P., Roy, E., Luminet, O., Fillée, C., & Mikolajczak, M. (2008). Relationship between alexithymia, alexithymia factors and salivary cortisol in men exposed to a social stress test. Psychoneuroendocrinology, 33, 1160–1164. Dichter, G. S., Tomarken, A. J., & Baucom, B. R. (2002). Startle modulation before, during and after exposure to emotional stimuli. International Journal of Psychophysiology, 43, 191–196. Elzinga, B. M., Schmahl, C. G., Vermetten, E., van Dyck, R., & Bremner, J. (2003). Higher cortisol levels following exposure to traumatic reminders in abuserelated PTSD. Neuropsychopharmacology, 28, 1656–1665. Eysenck, M. W. (2004). Trait anxiety, repressors and cognitive biases. In J. Yiend (Ed.), Cognition, emotion and psychopathology: Theoretical, empirical and clinical directions. Cambridge: Cambridge University Press. Fabian, L. A., McGuire, L., Page, G. G., Goodin, B. R., Edwards, R. R., & Haythornthwaite, J. (2009). The association of the cortisol awakening
S. Walker et al. / Personality and Individual Differences 51 (2011) 123–127 response with experimental pain ratings. Psychoneuroendocrinology, 34, 1247–1251. Faul, F., Erdfelder, E., Lang, A. G., & Buchner, A. (2007). G*Power 3: A flexible statistical power analysis for the social, behavioural, and biomedical sciences. Behaviour Research Methods, 39, 175–191. Friedman, H. S., & Booth-Kewley, S. (1987). The ‘disease-prone personality’: A metaanalytic view of the construct.. American Psychologist, 42, 539–555. Fries, E., Dettenborn, L., & Kirschbaum, C. (2009). The cortisol awakening response (CAR): Facts and future directions. International Journal of Psychophysiology, 72, 67–73. Infrasca, R. (1997). Alexithymia, neurovegetative arousal and neuroticism: An experimental study. Psychotherapy and Psychosomatics, 66, 276–280. Jezova, D., Makatsori, A., Duncko, R., Moncek, F., & Jakubek, M. (2004). High trait anxiety in healthy subjects is associated with low neuroendocrine activity during psychosocial stress. Progress Neuropsychopharmacology & Biological Psychiatry, 28, 1331–1336. Kenny, D. A., Kashy, D. A., & Bolger, N. (1998). Data analysis in social psychology. In D. T. Gilbert, S. T. Fiske, & G. Lindzey (Eds.), The handbook of social psychology. Boston, Massachusetts: McGraw-Hill Companies. Kudielka, B. M., & Kirschbaum, C. (2003). Awakening cortisol responses are influenced by health status and awakening time but not by menstrual cycle phase. Psychoneuroendocrinology, 28, 35–47. Kupper, N., de Geus, E. J., van den Berg, M., Kirschbaum, C., Boomsma, D. I., & Willemsen, G. (2005). Familial influences on basal salivary cortisol in an adult population. Psychoneuroendocrinology, 30, 857–868. Lang, P. J., Bradley, M. M., & Cuthbert, B. N. (1999). International affective picture system (IAPS): Instruction manual and affective ratings. Technical report A-4. The Center for Research in Psychophysiology, University of Florida, Gainesville, FL. Luteijn, F., & Bouman, T. K. (1988). The concepts of depression, anxiety, and neuroticism in questionnaires. European Journal of Personality, 2, 113–120. Marteau, T. M., & Bekker, H. (1992). The development of a six-item short-form of the state scale of the Spielberger state-trait anxiety inventory (STAI). British Journal of Clinical Psychology, 31, 301–306. McEwen, B. S. (2007). Physiology and neurobiology of stress and adaptation: Central role of the brain. Physiological Reviews, 87, 873–904. Miller, G. E., & Chen, E. (2006). Life stress and diminished expression of genes encoding the glucocorticoid receptor and b2-adrenergic receptor in children with asthma. Proceedings of the National Academy of Sciences of the United States of America, 103, 5496–5501. Newman, E., O’Connor, D. B., & Conner, M. T. (2007). Daily hassles and eating behaviour: The role of cortisol reactivity status. Psychoneuroendocrinology, 32, 125–132. O’Connor, D. B., Hendrickx, H., Dadd, T., Elliman, T., Willis, T. A., Talbot, D., et al. (2009a). Cortisol awakening rise in middle-aged women in relation to psychological stress. Psychoneuroendocrinology, 34, 1486–1494. O’Connor, D. B., Conner, M., Jones, F., McMillan, B., & Ferguson, E. (2009b). Exploring the benefits of conscientiousness: An investigation of daily stressors and health behaviors. Annals of Behavioral Medicine, 37, 184–196.
127
O’Connor, D. B., Jones, F., Conner, M., McMillan, B., & Ferguson, E. (2008). Effects of daily hassles and eating style on eating behaviour. Health Psychology, 27, S20–S31. Portella, M. J., Harmer, C. J., Flint, J., Cowen, P., & Goodwin, G. M. (2005). Enhanced early morning salivary cortisol in neuroticism. American Journal of Psychiatry, 162, 807–809. Preacher, K. L., & Leonardelli, G. J. (2001). Calculation for the Sobel test: An interactive calculation tool for mediation tests [Computer software]. Available from: