روان رنجوری و کورتیزول:برداشتن اثر مورد انتظار

There are strong theoretical arguments that those high on Neuroticism (N) should normally exhibit higher prevailing levels of the stress-linked hormone cortisol (C), but findings are inconsistent, probably reflecting methodological weaknesses especially in taking account of C's diurnal cycle.

According to the “wear and tear” model of McEwen and Wingfield (2010), some individuals have to work harder (i.e., expend more energy) to maintain homeostasis. The accumulation of this cost (allostatic load) implies a gradual decrease in the ability to respond to daily stressors. Hypothalamic–pituitary–adrenal (HPA) axis dysregulation (blunted cortisol responses to stress and high background levels of cortisol during the day) can be a consequence of such accumulated load. Cortisol, the HPA end product, is a glucocorticoid that our body naturally secretes according to a pronounced diurnal cycle with increased values evident under particularly stressful conditions. The typical diurnal profile shows a sharp rise upon awakening, which peaks 20–45 min later, and is called the ‘Cortisol Awakening Response’ (CAR). Thereafter there is a steady decline over the rest of the day with lower levels in the evening and night. A derivate measure, called area under the curve (AUC), is often used to estimate total secretion during a predefined time period (Hansen et al., 2008). Neuroticism (N), a relatively stable trait, is the predisposition to respond with intense emotional reactions to psychosocial stressors (Lahey, 2009). Individuals high on Neuroticism perceive and have more stressors, respond exaggeratedly to them, and require more time to recover from them (Suls and Martin, 2005). It has generally been assumed therefore that individuals high in N will tend to have an increased magnitude of cortisol secretion during the day, reflecting greater frequency and intensity of HPA stimulation from the psychosocial domain. A vast swathe of published theoretical statements on both the nature of Neuroticism and cortisol as a stress-responsive hormone appears to underpin such an assumption. It would be very puzzling therefore if the assumption were not at least broadly true, with the usual caveats of other things being equal etc. However, empirical studies, which directly or indirectly inform such a simple but highly plausible theoretical prediction, have been surprisingly inconsistent. Some findings (e.g. Nater et al., 2010) have been positive: higher mean daily AUC levels of cortisol were related to higher cortisol in the adult working population. Individuals with extremely high N scores were found to have greater waking cortisol levels compared to those with extremely low N scores (Portella et al., 2005). High N has been associated with significantly less flexibility of cortisol responding (though not over all cortisol level) when weekend and weekday profiles are compared (Mikolajczak et al., 2010). Flatter diurnal cortisol slope has been linked to higher N but only in male participants (Hauner et al., 2008). However in a majority of studies the predicted link between high N and high basal cortisol has eluded statistical significance (Chan et al., 2007, Ferguson, 2008, Gerritsen et al., 2009, Hill et al., 2013, Riese et al., 2009 and Van Santen et al., 2011). Finally, Neuroticism was not associated with differences in CAR magnitude in the meta-analyses done by Chida and Steptoe (2009). Inconsistency of effects does seem to be a prominent feature of the cortisol psychophysiology literature in general and might be explained by a host of methodological considerations. There are the usual problems in looking for consistency across studies. These can arise from diversity of sample populations, e.g. in regard to major demographics such as age and sex. They may be a product of differences in scales used to measure N as well as choice of continuous scale measurement or comparison of extreme groups. They arise from differing statistical power reflecting diversity in sample sizes, since associations of any kind with one of the broadest of all trait personality measures are likely to involve small effect sizes. However what is often less explicit but potentially crucial for cortisol measurement are factors relating to adequate timing and frequency of salivary cortisol sampling across the day, synchronization (or not) of cortisol sampling times to awakening times, and number of sampling days over which daily levels are averaged to arrive at supposedly valid ‘typical’ values over time, and last (but by no means least) adequacy of procedures to ensure participants' accurate adherence to protocol required sampling times, if saliva collection is carried out by participants. The present study was part of a larger research project with the overall aim of examining individual differences and various interventions on diurnal cortisol profiles of university students. In the present study, we sought sufficient methodological rigor to yield robust results for this population concerning theoretically expected associations between N and diurnal cortisol secretion. To do this, we sought to demonstrate replicability of effects over time, and given its pronounced diurnal cycle, we utilized objective checks on the timings of all cortisol samples. Therefore, over several year cohorts of students, we recruited samples with extreme high versus low scores for N (using the NEO-FFI measure), sampled from the diurnal cortisol cycle over two days rather than a single day, and verified timings of samples using MEMS (Track Caps) electronic monitoring device. We had one simple but clear theoretically derived formal hypothesis, namely that high N participants would ceteris paribus display elevated diurnal background levels of cortisol compared to low N participants. We also proposed to examine associations between N and the two pronounced dynamic aspects of the typical diurnal cortisol profile, namely the CAR and the subsequent diurnal fall over the course of the day. In respect of diurnal fall, we had an expectation, that high N might be associated with a less steep (i.e. ‘flatter’) fall, based on the latter's reported links to a number of stress-associated variables. In respect of the CAR, we did not consider the existing literature to permit of any firm a priori expectation in relation to existence or direction of any effect.