Chronic Stress in Cognitive Processes: Cortisol Dynamic Range of Secretion Is Associated with Perception of Unsafety Environment in a Venezuelan Population

Abstract

Background:

Understanding diurnal secretion of cortisol in association with behavioral attitudes as a result of perception of unsafety environment is a main interest in prospective studies establishing the impact of chronic stress in cognitive processes. Adaptive secretion of cortisol, a biomarker of the hypothalamic-hypophysis-adrenal (HPA) axis, has been correlated with perception of uncertainty in surroundings as a consequence of perseverative cognition and unconscious thoughts.

Objective:

To determine whether diurnal secretion pattern of cortisol was associated with behavioral attitudes indexes generated from answers to standardized questionnaires from Panamerican Health Organization/World Health Organization (PAHO/WHO) agencies.

Methods:

Saliva cortisol dynamic range was evaluated by immuno-essay. Cortisol awakening response (CAR) and total secreted cortisol was established in a cross-sectional study of four saliva samples per day from volunteers (n = 135) between 19 and 65 years old.

Results:

Saliva cortisol dynamic range followed a significant decay along the day. Reduction of social interaction and increase of defensive behavioral attitude was associated with older groups of age. In this study, two subgroups of subjects with a steeper cortisol secretion (slope significant non-zero), and flatter cortisol secretion (slope no significant non-zero) were detected. Noticeable, we determined an association between measurements of cortisol secretion from subjects with a flatter cortisol dynamic range and behavioral defensive and inhibition of social interaction indexes.

Conclusion:

These findings suggested chronical dysregulation of HPA axis as a result of perseverative cognitive perception of unsafety environment which may be precedent to cognitive impairment in the population.

Keywords

Alzheimer’s disease behavioral attitude chronic stress cognition cortisol mild cognitive impairment perception ruminant thoughts

INTRODUCTION

Classically, the stress response has been recognized as behavior repertoire of an individual when there is uncertainty about what to do to safeguard physical, mental, or social wellbeing. Indeed, uncertainty may concern either subject’s internal body milieu (lack of glucose, ATP or oxygen, loss of blood, infection or toxin challenging, trauma) or external surroundings (working environment, disordered or chaotic neighborhood, discrimination, social defeat) [1].

Stressors are responsible for either short term or prolonged physiological activity of the hypo-thalamic-pituitary-adrenal (HPA) axis modulating or over imposing an otherwise harmonized circadian dynamic response of cortisol secretion [2 –5].

The HPA axis hormones are widely thought to play a critical role in producing various detrimental outcomes of uncontrollable stress on brain structures. In fact, the prefrontal cortex and hippocampus are deemed sensitive, as a result of expression of high concentration of receptors, specifically corticosteroids (glucocorticoids and mineralocorticoids) in the hippocampus whose synthesis and secretion by the adrenal cortex are augmented by stress [6 –13].

Previous reports regarding range of cortisol secretion along several days and through the years has noticeable shown that aged humans with prolonged cortisol elevations showed reduced hippocampal volume and deficits in hippocampus-dependent memory tasks compared to normal-cortisol age matched controls [13 –16]. Altered HPA axis physiology, and in particular high cortisol levels in the elderly have been associated with an increased risk for dementia and Alzheimer’s disease (AD) [13 , 16–19].

Recently, in a seminal article, Brosschot et al. proposed stress function as a phylogenetically old pathway that is always set to high alertness, and is otherwise restraint as long as a safety environment can be perceived [20]. As a consequence, a behavioral character maintaining “On” the stress response when perception of uncertainty or unsafety is the rule would act as a strategy of survival with a benefit/cost advantageous by natural selection [20, 21].

The stress response is a physiological adaptive mechanism working “On” by default, which is modulated and inhibited or turned “Off” by prefrontal cortex activity when safety or certainty information from environment is dominant [21]. As a consequence, a balance between suppression of amygdala by cortical function and subsequent inhibition of behavioral stress-induced response (perception of safety) occurred until “unsafety” information is perceived again and, in a feed forward circuit, the cortical suppressive activity is reduced and the amygdala function take over the system [22 –24].

Consistent with this proposal, removal of inhibition from prefrontal cortex would “release” rather than “elicit” the stress response. Indeed, the supporting mechanism involved in the stress response is based in the concept of inhibition/disinhibition ratio which seems to be very common in biological systems [25], and has been known as the Hughlings Jackson principle of “hierarchical integration through inhibition” [26].

Prolonged stress-associated-behavior may occur as a result of anticipatory perception, recurrent cognition and slow recovery from representing stressors. Currently, abundant evidence supports that perseverative cognition of stressors affects cardiovascular, autonomic, and endocrine nervous system activity [21]. Commonly, the HPA axis is a crucial pathogenic pathway from chronic stress condition to long-term disease outcomes [13 , 27–31].

Coherent with this theory, is possible that either perception of unsafety surroundings in absence of a specific stress stimulus or anticipatory and perseverative perception of unsafety may elicit sustained stress response with potential detrimental effects in the organism such as sleep disorders, anxiety disorders and depression [32 –35].

In older adults, a lower cortisol awakening res-ponse (CAR) has been reported in those with generalized anxiety disorder [36]. In addition, greater diurnal cortisol secretion has been associated with frailty [37], whereas lower diurnal cortisol is correlated with longevity [38]. More specifically, altered morning-cortisol levels [39], or a flatter morning-to-evening diurnal cortisol slope may signify HPA axis dysregulation [27, 40]. Furthermore, subjects who perceive their neighborhoods as unsafe may express age-induced changes in HPA axis activity, such as increased diurnal cortisol secretion [41, 42].

There are behavioral, synaptic plasticity, and neu-ral activity data from animal studies which demon-strated that dynamic interaction between the prefrontal cortex, a structure implicated in the major hierarchical control of stressor [22 , 23–26], the amygdala, a major modulator of the stress response, and the hippocampus may modulate memory process associated to object recognition [10].

A better understanding of these interrelationships between cortisol, cognition, and dementia may open the door to new prevention and therapeutic options involving the HPA axis. The effects of cortisol on emotional memory had already led to therapeutic trials of corticosteroids and corticosteroid receptor antagonists/modulators in AD [13, 43].

Chronic stress and high cortisol levels may lead to an enhanced expression of pathologic processes of AD. For instance, in animal models of AD, administration of dexamethasone aggravates formation of cerebral amyloid plaques [44], and reduces clearance of cerebral amyloid [45]. Additionally, chronic stress and glucocorticoid administration induces hyperphosphorylation of tau, a key metabolic step in the formation of neurofibrillary tangles [46].

Studies in humans have shown that high corticosteroid levels may be detrimental to disease progression and AD risk. Findings of elevated concentration of cortisol in the urine has been found to be predictive of an increased risk for AD on average of 6 years before AD onset in a sample of cognitively healthy individuals from the Baltimore Longitudinal Study of Aging [18]. In addition, elevated cortisol levels have been found associated with faster cognitive decline of mild cognitive impairment patients and AD dementia [47, 48].

The goal of this study was to investigate whether daily range of cortisol secretion was associated with psychosocial and behavioral attitudes as a result of a perception of unsafety environment. For that purpose, we have established: 1) a representative sampling of a population carefully selected to ensure that participants reflect the characteristics of semi-urban population of interest; 2) a large sample of the population with a broader socioeconomic status; and 3) extensive measurement of cortisol dependent variable.

We used a standardized questionnaire in reference to: 1) preconceived thoughts, 2) perception of unsafety environment, and 3) inhibition of social interaction as mechanism of self-preservation. By analyzing answers of volunteers, we generated behavioral indexes for the complete sampled population and correlated with diurnal range of measurements of cortisol secretion. A further objective was to extend the biological value of cortisol by dissecting comprehensively secretion of the hormone as sub-sampled dependent variable with respect to time along the day and behavioral attitude of respondents.

Taken as a whole, this study evaluates cortisol daily time course, cortisol awakening response, and cortisol decay response, altogether associated with behavioral attitudes as a result of perception of unsafety from environment.

METHODS

Participants (N = 135) were voluntarily added to the protocol in three visits to Valle la Pascua city (Municipality of Leonardo Infante) in the Guárico state in the Center North Eastern area of the mainland of the Bolivarian Republic of Venezuela (Fig. 1). This location has a mestizo population of 115,902 inhabitants (Statistical National Institute, 2011), consisting of farmers and livestock breeders.

Fig. 1

A) Map of Venezuela in South America. B) Political map of Venezuela with the study area (Guarico State in grey; star represent relative location of Valle la Pascua.

Collection and analysis of data were approved by the Bioethical committee of the University Hospital of Caracas on behalf of the Consejo de Desarrollo Cientifico y Humanístico of University Central of Venezuela (CDCH-UCV, Caracas, Venezuela), and the Institute of Pharmaceutical Investigation, School of Pharmacy, University Central of Venezuela (IIF-FF-UCV).

All studies were conducted in compliance with the Declaration of Helsinki, and all participants provided informed consent. The current study included participants with full information on multiple measures of cortisol and key outcomes, including psychosocial factors, chronic medical conditions, and socio-demo-graphic factors.

Participants were enrolled voluntarily in the protocol through a broadcasting public request from Valle la Pascua city radio station which publicized the study. As a result of the social media calling a heterogeneous population was incorporated to the protocol with a larger proportion of volunteers from the Valle la Pascua Technological Institute. We observed no differences in the mean age between females (41.91±3.34) and males (48.23±1.51) volunteers.

All individuals were asked to answer a standard-ized questionnaire of the Division of non-transmis-sible diseases of the Panamerican Health Organization (PAHO) regional office from World Health Organization (WHO). All volunteers received a standardized questionnaire involving multiple items to assess socio-economic status, health habits (tobacco, alcohol, and food consumption), and chronic diseases (hypertension, diabetes).

Participants rated themselves where they stood in society in terms of education, occupation, and wealth. Socio-economic status of the sample population was established taking into account level of literacy which was constituted mainly by high school educated and post-secondary graduates. The majority of participants did not smoke (63%). With regards to some health characteristics we found in the sampled population, 54% were overweight 54% , 38% had hypercholesterolemia, and 85% had hyperglycemia (Table 1).

Table 1

Demographics of the population from Valle la Pascua (Guarico, Venezuela)

Age (M, SD)	46.52±4.24
Female (%)	30
Education
University (%)	55
High School/Some College (%)	33
Working (%)	65
Home Ownership (%)	64
Health Characteristics
Health Insurance Coverage (%)	58.33
Visits to Physician (last 12 months)	81
Avoiding consultation because of high price	22
Chronic Conditions (%)
Currently smoking cigarettes (%)	32
Overweight (%)	54
High Cholesterol (%)	37
High Glycemia (%)	84
High blood pressure (%)	16

Quantification of cortisol in saliva samples

Each volunteer received a detailed explanation of the procedure for collecting the saliva sample for cortisol measurements. All participants received an individual kit consisting of plastic bag with a logo data sheet paper with indications of the sampling procedure. In addition, four micro-centrifuge tubes (1.5 ml) filled with a piece of dental cotton roll and a disposable eyebrow tweezer to manipulate pieces of cotton were also included in the plastic bag.

Each subject itself was trained to collect 4 saliva samples: at time of awakening, 2 hours later, at noon and at 6 pm before dinner time. At the end, they were asked to maintain all samples at 4°C overnight until return the plastic bag kit to the collector. Volunteers that failed in collecting the complete set of samples were excluded from the analysis.

Participants were informed about the need to adhere to the strict sampling regime relative to awakening time and accuracy of their saliva sampling. They were also instructed to awake in their usual way either spontaneously or by an alarm clock. During the saliva collection period participants were indicated to refrain from smoking, brushing their teeth, exercising, and taking anything by mouth except water. Samples were returned to the laboratory to be stored at –20°C until assayed. Participants were asked to fill in a record sheet on each day entering their awakening times, and their actual saliva sampling times.

Cortisol concentrations were determined by DSL-10-671000 ACTIVE®, an enzyme linked immuno-sorbent assay developed by Salimetrics (USA) for cortisol secreted in the saliva. Treatment of samples was as follows, shortly: all specimens and reagents were kept to room temperature (25°C) during the analysis. Standards, controls, and unknowns were assayed in duplicate.

Procedures were performed as described in the package insert: 25μl of each standard, control, and unknown samples were dispensed into the appropriate wells. Then, 100μl of the enzyme conjugate solution were added to each well. After gently tapping the well holder for 5–10 s, 100μl of the cortisol antiserum was added to each well and incubated at room temperature (25°C) on a shaker set at 500–700 rpm for 45 min. Each well was aspirated and washed 5 times with the Wash Solution, then dried by inverting plate on absorbent material. Finally, 100μl of the TMB chromogen solution was dispensed to each well and incubated at RT (25°C) for 10–15 min on an orbital microplate shaker set at 500–700 rpm. Using a microplate reader set 450 nm, the absorbance of the solution in the wells was read within 30 min.

Establishing life habits and social behavior

By using the PAHO-WHO standardized questionnaire we established three environmentally-related behavioral attitudes which were selected to generate a relative distress index taking into account responses to the PAHO/WHO questionnaire from the Program of non-transmissible diseases.

Only 85 volunteers out of 135 total participants answered all sub-questionnaires; as a consequence, all additional behavioral analysis were performed with this group of subjects. Taking this standardized tool, detailed questions and phrases associated with each sub-class (Table 2A-C), three behavioral attitudes were considered: 1) a Behavioral Defensive Index (BDEFENSI), the lower the total value meaning more defensive attitude; 2) Behavioral Unsafe Index (BUNSAFI), the lower the total value meaning more safety; and 3) Behavioral Index of social interaction Inhibition (BINHIBI), the lower the total value means more behaviorally inhibited. All questionnaire measurements provided standardized sub scale scores on these dimensions of behavior.

Table 2A

Behavioral Defensive Index (BDEFENSI)

		FA	SA	NC	SD	FD
Do you agree or disagree with the following sentences?	You have the right to kill if your family is in danger
FA: Fully Agreed SA: Some Agreed	You have the right to kill to defend your home place and your property
NC: No comments SD: Some disagreed	If you have your own firearm, your home place is safer
FD: Fully Disagreed	Carrying a firearm makes you feel safer

Table 2B

Behavioral Unsafe Index (BUNSAFI)

		VS	SS	SU	VU
How do you feel in the following places?	In your house or apartment
VS: Very Safe SS: Some Safe	In streets around your neighborhood during the day
SU: Some Unsafe VU: Very Unsafe	In streets around your neighborhood during the evening
	In the public transportation
	In down town the city

Table 2C

Behavioral Inhibition Index (BINHBI)

		V	S	N
Because you are afraid of being victimized with violence, currently:	You have limited number of places for shopping
V: Very S: Some N: Nothing	You have limited number of recreation activities than before
	You feel compelled to carry on a weapon for self defense
	You feel compelled that moving to another city or another country

All volunteers were asked to read the informed consent, fill the questionnaire, and if they agreed, a blood sample (3 ml) was taken by a dedicated phlebotomist from the cubital vein, and reserved at 4°C until DNA extraction. Participants received no financial incentive to take part in the study.

Determining apolipoprotein E genotype

For the determination of APOE polymorphisms, restriction enzyme isoform genotyping using the PCR of DNA amplification methodology was performed [49]. Leukocyte DNA was extracted from subjects who were previously informed with respect to APOE isoforms and the meaning as risk factor of AD pathology. Leukocyte DNA was amplified by PCR in a DNA Thermal Cycler (BioRad, USA) using oligonucleotide primers: F4 (5^′-ACAGAATTCGCCCCGGCCTGGTACAC-3^′) and F6 (5^′TAAGCTTGGCACGGCTGTCCAAGGA-3^′).

In addition to the buffer and nucleotide components described by the supplier of Taq polymerase (Invitrogen) each amplification reaction contained 1μg of leukocyte DNA, 1 pmol/μl of each primer, 10% dimetil sulfoxide, 0.025 units/μl of Taq polymerase in a final volume of 30μl. Each reaction mixture was heated at 95°C for 5 min for denaturation, and subjected to 30–35 cycles of amplification by primer annealing (60°C for 1 min), extension (70°C for 2 min), denaturation (95°C for 1 min), and a final extension (70°C for 10 min).

We obtained a 300 bp of DNA product of amplified APOE sequences. Amplification of APOE products that were suitable for HhaI digestion proved successful for most DNA samples, with the exception of samples that were extensively degraded prior to amplification. Restriction analysis of amplified APOE sequences was made with HhaI restriction enzyme and gel analysis after PCR amplification. In short: 5 units of HhaI (New England Biolabs) were added directly to each reaction mixture for digestion of APOE amplification products (>3 h at 37°C). This process did not require purification of PCR products or addition of specific buffer components for HhaI digestion.

Each reaction mixture was loaded onto an 8% polyacrylamide non-denaturing gel (1.5 mm thick x 25 cm long) and electrophoresed for 3 h under constant current (45 mA). After electrophoresis, the gel was treated with ethidium bromide (0.2 mg/l) for 10 min and DNA fragments were visualized by UV illumination. The sizes of HhaI fragments were estimated by comparison with known size 20 bp DNA standardized DNA marker.

Statistical analyses

Assessment of cortisol secretion: area under curve/period (AUC6-12 h) and AUC12-18 h, AUC6-8 h or cortisol awakening response (CAR).

Cortisol concentrations ranged between 0.004 and 55 nmol/l and values were moderately positively skewed. To normalize sample distributions for inferential analyses, log transformation was performed. AUC was calculated from polygon considering peak to ground which is assumed to be a measure regarding total hormonal output [50, 51].

CAR was determined as the AUC6-8 between the first sample at awakening time declared by the volunteer and the second sample 2 h later. AUC was calculated from the secreted cortisol at each time point along the day per subject. AUC from either the determined cortisol at each time point and correspondent transformed values were used for calculating the fraction of secretion for three defined intervals of time AUC6-8 h or CAR, AUC6-12 h, and AUC12–18 h. An analysis of variance (ANOVA) of repeated measures was carried out to compare the values for cortisol along the day, and an interaction analysis post-hoc was carried out to further analyze (post hoc Tukey tests).

A general linear model of regression analysis was used to adjust the secreted cortisol values in the main diurnal time courses of 12 h. The coefficient of regression was taken into consideration to discriminate between curves with a significant decay (Regression coefficient significant non-zero at p < 0.05 and no significant non-zero (p > 0.05). Using this approach, we detected two sub-populations as follows: a subpopulation with a regression coefficient significant non-zero, it means steeper cortisol decay along the day, and another with a regression coefficient no significant non-zero, e.g., flatter decay of cortisol secretion.

Correlation analysis were performed between cortisol calculated measurements and indexes generated from each individual’s interview. 85 volunteers answered all questionnaires and their saliva set of samples were collected. Values from each of the three sub-questionnaires were used to calculate Pearson correlation coefficient: 1) BDEFENSI, 2) BUNSAFI, and 3) BINHIBI with respect to age, AUC6–18 h, AUC12–18 h, AUC(ln)12–18 h, and AUC12–18 h “flatter curve of cortisol”. In addition, Chi square analyses (χ²) were performed to discriminate between homogenous versus heterogeneous frequency of responses asked under “violence” sub questionnaire. All analyses were considered significant when p < 0.05, and were conducted using GraphPad Prism version 6.00 for Windows, GraphPad Software, La Jolla California USA, http://www.graphpad.com.

RESULTS

Establishing a daily pattern of cortisol secretion

Assessment of cortisol secretion in saliva samples showed a diurnal pattern that was generated from our four sample protocol that started with a highest value at awakening time (22.02±0.90 nmoles/l), and a significant reduction two hours later (16.34±0.60 nmoles/l). Then, subsequent lower values at noon (11.82±0.45 nmoles/l), and minimum value around 6 pm (6.92±0.42 nmoles/l). Self-collection of saliva samples allowed us to determine a time course of cortisol secretion with a significant decay along the day (ANOVA: F(3; 519) = 104.6; p < 0.0001). An individual post-hoc analysis of concentrations of cortisol at each time point also demonstrated a significant reduction at p < 0.05 level (Fig. 2).

Fig. 2

Daily time course of cortisol in the Valle la Pascua sampled population. Each bar represents group values of at least 4 samples per subject as described in Methods. Symbols are cortisol concentration per subject determined by ELISA methodology. Data are the mean±SEM of N = 124 to 135 subjects. a) Letters on top lines indicate significance: a) One way ANOVA F (3,519) = 104.6; p < 0.0001. b) p < 0.05 between groups.

There was no gender difference in secretion of cortisol at the awakening time point or at any other time between females and males (results not shown).

We assessed four saliva samples per individual to calculate the total amount of cortisol secreted along the day, or area under curve (AUC_6–18 h) for each participant. The mean value of cortisol or AUC_6–18 h was 152.97±4.04 nmoles/l (n = 139) for the complete sampled population. Additionally, analysis of the mean AUC_6–18 h value between females and males, showed no differences. As a consequence, the AUC_6-18 h value of cortisol released for each gender was also no different (results not shown).

We determined that the AUC_6–18 h values between groups organized by decades of age, starting at 20 and until 60 years old, were not different (results not shown). The 40 to 60 years old group of participants in this study was larger, representing around 60% of total sample.

Cortisol secretion in the saliva samples followed a time-dependent function that may be described by two linear components, as follows: a time orderly component of 6 h between the awakening time, around 6 am until noon with an AUC_6–12 h value of 96.18±3.09 ng/ml, followed by a second component between noon and 6 pm with a AUC_12–18 h value of 58.70±2.48 ng/ml. That difference was highly significant (p < 0.0001, paired t test, Fig. 3).

Fig. 3

Total concentration of cortisol secreted estimated after area under curve (AUC) calculation from morning (AUC_6–12 h) and afternoon (AUC_12–18 h) time periods. Values are the mean±SEM of AUC_6–12 h) (N = 139) and (AUC_12–18 h) (N = 136) individuals. Top line represents significance in a non-paired t test (9.03; 273) = p < 0.0001.

A time-dependent decay in cortisol secretion was described by a typical non-linear regression when the whole daily time course was considered. Thus, a faster decay in cortisol secretion occurred during the morning hours relative to the afternoon hours. However, we noticed that an appropriate descriptor of the physiological decay of the cortisol secretion was a general model of linear regression. As a result of this analysis we observed a fraction of the population with a slope significant non-zero (p < 0.05) that was higher with respect to the subpopulation with a slope no significant non-zero (p > 0.05) (Fig. 4).

Fig. 4

Comparison of regression coefficient (slope, β) estimated from linear regression analysis (LRA) model from cortisol values at four time points as described in methods. Values are the mean±SEM of slopes from significant non-zero (N = 44) with respect to nonsignificant zero (N = 92) individuals. Top line represents significance in an unpaired t test (4.41; 168) = p < 0.0001.

A significant decay in cortisol secretion along the day was suggestive of harmonized modulation by the HPA axis, and when absolute measurements of cortisol of each participant were used to establish an individual function of regression, potential intrinsic properties of physiological response were revealed, as follows: First at all, we detected a subpopulation of participants with a coefficient of linear regression significant non-zero (N = 46), with a steeper cortisol function decay along the day, suggestive of function and as a consequence homeostatic regulated cortisol secretion by HPA axis. However, interesting and noticeable, there was another group of individuals with a coefficient of linear regression (slope) no significant non-zero, i.e., with a flatter function secretion of the hormone. Two additional subpopulations with a coefficient of linear regression (slope-) non-significant non-zero in an interval of probability (0.05 < p < 0.075; n = 14) and (0.075 < p < 0.10; n = 10), of 14 and 10 individuals, were also detected; these additional groups were included in the no significant non-zero group.

Further analyses were performed in both “coefficient regression significant non-zero”, and “coefficient regression no significant non-zero” with a flatter curve of cortisol secretion groups. The value of cortisol at waking time from the sub-population with a flatter curve of cortisol secretion, was lower with respect to the corresponding time point of the coefficient regression significant non-zero sub-group (p < 0.012), and higher at 18 h (p < 0.0001). Coherent with this observation, in the sub-population with a flatter curve of cortisol secretion, the CAR value calculated from the AUC6–8 h period of time was lower with respect to CAR of with the significant non-zero group (p < 0.0001) (Fig. 5A).

Fig. 5

Cortisol secretion in (A) CAR (AUC_6–8 h) and (B) AUC_12–18 h periods of time. Cortisol values at CAR and afternoon were calculated as described in methods from subjects with slopes significant non-zero with respect to non-significant zero. Total values of concentration of cortisol estimated after area under curve (AUC) calculation are the mean±SEM from significant non-zero (N = 44) with respect to nonsignificant zero (N = 92) groups of individuals. Top line represents significance in an A) unpaired t test (4.29; 137) = p < 0.0001 and B) (3.33; 135) = p < 0.0011.

Noticeable, the mean of AUC_12–18 h value of cortisol for the afternoon period was higher in the subpopulation with a flatter curve of cortisol secretion with respect to AUC_12–18 h from the coefficient regression significant non-zero group (Fig. 5B). As a consequence, these results were suggestive that cortisol secretion is heterogeneous in the sampled population, which was not evident when daily range of cortisol from the complete population was considered. In addition, these findings may also support a possible dysregulation of HPA axis in the no significant non-zero “flatter curve of cortisol secretion” group.

Cortisol dynamic range and perception of unsafety in surroundings

To ascertain if cortisol secretion is related to the perseverative cognition of stressors, we first examined whether there is an age-related environmentally-associated behavior attitude. For that purpose, a correlation matrix between age and all three behavioral attitudes indexes were generated. First, a significant correlation was found between AGE and BINHIBI (Table 3) suggesting that aged individuals spontaneously may limit social interactions, for example reducing recreation activities and going to open places to do shopping (r = –0.26; p < 0.015). Second, coherent with these results, subjects may possibly search for more secure areas. Indeed, a limited number of places for shopping was associated with age (BINHIBI.Q1 versus AGE, r = –0.22; p < 0.042). By contrast, no correlation was observed between age and neither BDEFENSI nor BUNSAFI, suggesting no association between age and either active behavioral defensive or inhibitory social interaction.

Table 3

Association between cortisol metrics and standardized Behavioral Attitude Indexes (Pearson Correlation Coefficient)

Variable	BDEFENSI				BINHIBI
	r = -	Low = -	High = -	p < 0.05	r = -	Low = -	High = -	p < 0.05
Age (N = 84)					0.26	0.05	0.45	0.015
					INHIBI. Q1
Age (N = 84)					0.22	0.008	0.41	0.042
AUC_12–18 h (N = 84)	0.24	0.029	0.43	0.026	0.24	0.02	0.43	0.026
	BDEFENSI.Q1				INHIBI. Q1
AUC_12–18 h (N = 84)	0.27	0.06	0.45	0.011	0.24	0.03	0.43	0.020
	BDEFENSI.Q2
AUC_12–18 h (N = 84)	0.25	0.03	0.43	0.02

Since, as shown here, the AUC_6–18 h of cortisol in the saliva results from at least two kinetically different linear components, we asked whether either morning secretion or afternoon secretion of cortisol may be correlated with BDEFENSI, BUNSAFI, or BINHIBI indexes.

We explored whether AUC_6–12 h of cortisol values were correlated with BDEFENSI, BUNSAFI, and BINHIBI indexes; and found no significant correlation. By contrast, when we examined the cortisol secretion represented by the AUC_12–18 h secretion (Table 3), significant correlations for BDEFENSI (r = –0.24; p < 0.026) and BINHIBI (r = –0.24; p < 0.026) were observed, but not with BUNSAFI. Taking together these results suggest an association between AUC_12–18 h cortisol secretion and these behavioral attitudes. In addition, afternoon secretion was associated with two behavioral attitudes of self-protection suggested by BDFENSI.Q1 (r = –0.27, p < 0.011) and BDFENSI.Q2 (r = –0.25, p < 0.02) as well as BINHIBI.Q1 (r = –0.24; p < 0.02).

Taking into consideration that a flatter curve of cortisol secretion may be indicative of a dysregulation of the HPA axis and associated to an unsafety perception of the environment, we explored the association between AUC of cortisol secretion with behavioral attitudes indexes from those subjects for whom the coefficient regression was no significant non-zero. The AUC_6–18 h of cortisol was associated with BINHIBI (r = –0.25; p < 0.054), BINHIBI.Q1 (r = –0.22; p < 0.042), and BINHIBI.Q2 (r = –0.27; p < 0.03) but neither BDEFENSI nor BUNSAFI association were significant (Table 4). We asked whether cortisol secretion during afternoon may be associated with perseverative cognitive defensive behavioral attitude. Interesting, when AUC_12–18 h of cortisol secretion and AUC(Ln)_12–18 h were considered with respect to answers associated to subcomponents of BDEFENSI and BINHIBI questionnaire index, a significant association was demonstrated for BDEFENSI.Q1 (r = –0.29; p < 0.047), BDEFENSI.Q2 (r = –0.32; p < 0.02), BDEFENSI.Q3 (r = –0.32; p < 0.025), and BINHIBI.Q1 (r = –0.32; p < 0.024). Taking together, these results suggest that daily dynamic secretion of cortisol is related to inhibitory or defensive behavioral attitudes.

Table 4

Association between “flatter cortisol dynamic range measurement” and standardized Behavioral Attitude Indexes (Pearson Correlation Coefficient)

Variable	BDEFENSI				BINHIBI
	r = -	Low = -	High = -	p < 0.05	r = -	Low = -	High = -	p < 0.05
AUC_6–18 h (N = 57)					0.25	0.005	0.48	0.054
					INHIBI. Q1
AUC_6–18 h (N = 57)					0.22	0.008	0.41	0.042
					INHIBI. Q4
AUC_6–18 h (N = 57)					0.27	0.01	0.49	0.03
	BDEFENSI.Q1				INHIBI. Q1
AUC_12–18 h (N = 47)	0.29	0.003	0.53	0.047	0.32	0.043	0.56	0.024
	BDEFENSI.Q2
AUC_12–18 h (N = 47)	0.32	0.043	0.56	0.02
	BDEFENSI.Q3
AUC_12–18 h (N = 47)	0.32	0.04	0.56	0.025
	BDEFENSI.Q4
AUC(Ln)_12–18 h (N = 45)	0.30	0.01	0.55	0.04

The “violence” sub-questionnaire may consider a perseverative cognitive agonistic behavioral attitude. We named “active” when some questions are suggestive of an active behavioral attitude associated with punishment, for example: “Do you agreed or fully disagreed with the following questions”: 1) “punish a child as a pedagogical strategy”, 2) “punish a woman that is unfaithful to her husband”, 3) “a woman has the right to punish another woman that is harassing her husband”, 4) “if the authorities fail the people, they are allowed to take the justice for themselves”.

We asked whether the selection of answers in these enquiries generated either a homogeneous or heterogeneous response from participants.

In regards of generalized perception of unsafety environment, we found that some volunteers were more confident “inside home” than outside in the following environments, as follows: streets, χ² (72.76, 3; p < 0.0001); streets at night, χ² (40.76, 3; p < 0.0001); public transportation, χ² (71.40, 3; p < 0.0001); down town, χ² (69.58, 3; p < 0.0001). In addition, they feel more confident in “streets” with respect to: streets at night; and public transportation, χ² (8.43, 3; p < 0.038) and down town χ² (69.58, 3; p < 0.0001).

Interestingly, some questions where a hypothetical condition may involve contrasting violent acts against child or women generated a heterogeneous response. Thus, we noticed that questions as: “punishing a child as “educative” strategy”, or “inflicting damage to a woman by her husband because she has been unfaithful”; generated a heterogeneous response χ² (8.98, 3; p < 0.0027) in the sampled population.

Determining the allelic frequency of APOE

The percentage of APOE alleles was: ɛ2 4.92% , ɛ3 83.82% , and ɛ4 11.27% . As a result, the observed frequency of all possible phenotypes E3/E3, E3/E4, E2/E3, E2/E4, and E4/E4 in the population were as follows: 74, 16, 7, 3, and 2, respectively. Following the analysis suggested by using the Hardy-Weinberg model, the observed frequency was not different from to expected frequency in the population studied (χ² = 4.59, 2; p > 0.05). No association was observed with the AUC_6-18 h of cortisol determined for the complete population.

DISCUSSION

In this study, assessment of cortisol along the day and time course analysis with a general model of regression suggested a heterogeneous profile of hormone secretion and modulation by the HPA axis. We have detected two groups in the studied population, one of them with a characteristic steeper and significant decay in cortisol secretion, and another group with a daily flatter and non-significant decay of hormone secretion.

Association of several measurements of cortisol secretion function with behavioral attitudes indexes demonstrated for the first time the impact of representation of stressful situations and unconscious thoughts on modulation of cortisol secretion by HPA axis. Following our experimental approach, self-collection of saliva at scheduled times provided some benefits, such as reduced stress-induced response by the collector. Even so, it is possible that lack of supervised collection of the saliva sample may also add uncertainty as a result of uncomplete adhesion to the time of collection and handling of the sample. However, reduced variability in the concentration of cortisol at each time point suggested that the methodological strategy in the proposed schedule of sampling collection was appropriate.

Cortisol dynamic range was studied taking into account well known reporters of HPA axis physiology. For instance, CAR value was estimated from AUC_6-8 h of cortisol at waking time and 2 h later. Some limitations have been pointed out in reference to the delay between waking time and collection of saliva to establish raising cortisol peak to determine CAR. Indeed, 10 to 30 min after waking time, a raise in the cortisol secretion occur and remains elevated for 60 min [50 –53]. Several reports have defined time limits for sampling to estimate CAR secretion until 3 h after awakening time [16 , 54–56]. Thus, collection of saliva samples in a time window of 2 h was suitable to obtain an unbiased measurement of a CAR of cortisol [57, 58]. In doing so, CAR value was reduced in the flatter cortisol dynamic range subgroup of the sampled population.

Our findings show that there are no changes in cortisol secretion associated with either aging or gender. These findings suggest that cortisol diurnal patterns are preserved during aging. These results are consistent with previous reports [53, 59]; however, prior studies have reported reduction [60 –62], or no association with age [63]. It is probable that differences in populations, assessment tools, as well as confounding factors including age, gender, educational level, and other neuroendocrine and psychological factors might explain these discrepancies.

In agreement with previous reports, analyses of single measurements along the day, and either CAR or fractional cortisol slopes suggested subtle differences in individual physiological response with psychophysiological prospective value [27]. Indeed, several studies have reported that cortisol secretion reaches a peak within an hour of waking and declines thereafter. A consistent CAR followed by a steep decline of cortisol secretion has been thought to reflect healthy HPA axis function [27, 52].

Following several cortisol measurements such as CAR or AUC_6-8 h, daily AUC_6-18 h, AUC_6-12 h (morning), and AUC_12-18 h (afternoon) in the same population, we have been able to determine individual differences in the cortisol function with a common pattern occurring along the day. In agreement with previous reports, evaluating cortisol dynamic range as function reporter associated with behavioral measurements may help understanding the effects of psychosocial stressors on health and disease [13 , 64–66].

Further observation of cortisol daily secretion of the population was suggestive of at least two subgroups according to either coefficient regression significant non-zero or no significant non-zero (p < 0.05) (slope) analysis. From these data, 54% of the sample subjects showed steeper decreasing pattern of cortisol, whereas 46% showed a flatter cortisol secretion along the day with a slope no significant non-zero. Interestingly, these findings were unexpected considering that a significant reduction of cortisol at each time point was found by ANOVA for the entire population as we showed in results.

Noticeable, presence of flatter diurnal cycle could possibly be indicative of a dysregulation of the HPA axis due to intrinsic conditions or perceptive altered environmental circumstances as we determined by correlation analysis. From these findings, a sustained stress response may occur as a result of either recurrent perception of stressor stimuli or anticipated perception of unsafety condition rather than stressors themselves [67 –69]. Clearly, even though a daily time course in secretion of cortisol remains under HPA axis activity, sensitivity of cortisol secretion control system was altered in the subgroup with a slope no significant non-zero. This is the first demonstration, to the best of our knowledge, of an association between cortisol dynamic range and behavioral attitude indexes as a measurement of unconscious stress.

The stress response mediated by subcortical limbic structures, including the amygdala, appears to be normally under continuous inhibition by the ventromedial prefrontal cortex [24 , 71]. When signals of certainty or safety are relevant, subcortical inhibition by the prefrontal cortex appears to be enhanced and amygdala activity is reduced. Conversely, recurrent thoughts of uncertainty and unsafety or perseverative cognition would remain even after a stressor has stopped, when perception of safety is still in question [72], and chronically even though no threats are present [73, 74].

Findings of this study were suggestive of subgroup of subjects that seems to be associated to a dysfunctional HPA axis in a cognitively normal population. Several reports have pointed out either dysregulated cortisol or personality traits as neuroticism may play a role as risk factor of mild cognitive impairment and AD [30].

Altered HPA axis functioning, and in particular high cortisol levels, in the elderly have been associated with an increased risk for dementia and AD [16 –19]. For instance, glucocorticoids increased potency of amyloid-β on oxidative injury and as a result toxicity in cultured hippocampal neurons [75]. Even more, high cortisol levels have been associated with tau and amyloid-β peptide pathology in a mouse model of AD [76].

Recent reports have described an increase in expression of cortisol in cerebrospinal fluid [44, 48], and in mild cognitive impairment due to AD, high cerebrospinal fluid cortisol was also predictive of a more rapid cognitive decline [48].

In summary, a homogenous behavioral attitude such as perceptive unsafety producing defensive behavior or inhibition of social interaction has some advantages and greater explanatory ability for prolonged and even chronic stress responses, when either occur cognitive representations of stress stimuli or when no actual stressors are present at all.

Altogether findings from this study, a flatter dy-namic diurnal course of cortisol was suggestive of a susceptible population with a dysfunctional HPA axis with a potential of medium to long term impairment of cognitive process as has been proposed in previous reports [12 , 77–79].

For epidemiological purposes this study validates cortisol measurements in the general population as a biomarker that would reinforce social and health intervention in sensitive areas where environmental conditions are unsafe.

Footnotes

ACKNOWLEDGMENTS

This work was supported by CDCH PG0610-5379-05 from Consejo de Desarrollo Científico y Humanístico de la Universidad Central de Vene-zuela and FONACIT MPPCT Science Mission N° 200701493.

The authors acknowledge S. Ruiz from the Instituto Universitario Tecnológico de Valle la Pascua, Estado Guárico, Venezuela, and J. Villegas, R. Pan-Dávila, and L. Martinez de Bracho from the Hospital Rafael Zamora Arèvalo, Valle la Pascua, Estado Guárico, Venezuela.

Authors’ disclosures available online (). We appreciated the graphical work of maps from Maria Fernanda Vicenti.

References

Peters

, McEwen

, Friston

(2017) Uncertainty and stress: Why it causes diseases and how it is mastered by the brain. Prog Neurobiol156, 164–188.

Adam

, Quinn

, Tavernier

, McQuillan

, Dahlke

, Gilbert

(2017) Diurnal cortisol slopes and mental and physical health outcomes: A systematic review and meta-analysis. Psychoneuroendocrinology83, 25–41.

Sapolsky

, Krey

, McEwen

(1986) The neuroendocrinology of stress and aging: The glucocorticoid cascade hypothesis. Endocrine Rev7, 284–301.

Pruessner

, Wolf

, Hellhammer

, Buske-Kirschbaum

, von Auer

, Jobst

, Kaspers

, Kirschbaum

(1997) Free cortisol levels after awakening: A reliable biological marker for the assessment of adrenocortical activity. Life Sci61, 2539–2549.

Lightman

(2008) The neuroendocrinology of stress: A never ending story. J Neuroendocrinol20, 880–884.

McEwen

, de Leon

, Lupien

, Meaney

(1999) Corticosteroids, the aging brain and cognition. Trends Endocrinol Metab10, 92–96.

McEwen

, Morrison

(2013) The brain on stress: Vulnerability and plasticity of the prefrontal cortex over the life course. Neuron79, 16–29.

McEwen

(2007) Physiology and neurobiology of stress and adaptation: Central role of the brain. Physiol Rev87, 873–904.

McEwen

(1998) Stress, adaptation, and disease: Allostasis and allostatic load. Ann N Y Acad Sci840, 33–44.

10.

Kim

, Kim

(2019) Amygdala, medial prefrontal cortex and glucocorticoid interactions produce stress-like effects on memory. Front Behav Neurosci13, 210.

11.

Kim

, Haller

(2007) Glucocorticoid hyper- and hypofunction: Stress effects on cognition and aggression. Ann N Y Acad Sci1113, 291–303.

12.

Gardner

, Lightman

, Kuh

, Comijs

, Deeg

, Gallacher

, Geoffroy

, Kivimaki

, Kumari

, Power

, Hardy

, Richards

, Ben-Shlomo

(2019) Dysregulation of the hypothalamic pituitary adrenal (HPA) axis and cognitive capability at older ages: Individual participant meta-analysis of five cohorts. Sci Rep9, 4555.

13.

Ouanes

, Popp

(2019) High cortisol and the risk of dementia and Alzheimer’s disease: A review of the literature. Front Aging Neurosci11, 43.

14.

Lupien

, Lecours

, Schwartz

, Sharma

, Hauger

, Meaney

, Nair

(1996) Longitudinal study of basal cortisol levels in healthy elderly subjects: Evidence for subgroups. Neurobiol Aging17, 95–105.

15.

Lupien

, de Leon

, de Santi

, Convit

, Tarshish

, Nair

, Thakur

, McEwen

, Hauger

, Meaney

(1998) Cortisol levels during human aging predict hippocampal atrophy and memory. Nat Neurosci1, 69–73.

16.

Lupien

, Nair

, Brière

, Maheu

, Tu

, Lemay

, McEwen

, Meaney

(1999) Increased cortisol levels and impaired cognition in human aging: Implication for depression and dementia in later life. Rev Neurosci10, 117–139.

17.

Rothman

, Mattson

(2010) Adverse stress, hippocampal networks, and Alzheimer’s disease. Neuromolecular Med12, 56–70.

18.

Ennis

, An

, Resnick

, Ferrucci

, O’Brien

, Moffat

(2017) Long-term cortisol measures predict Alzheimer disease risk. Neurology88, 371–378.

19.

Notarianni

(2017) Cortisol: Mediator of association between Alzheimer’s disease and diabetes mellitus?Psychoneuroendocrinology81, 129–137.

20.

Brosschot

, Verkuil

, Thayer

(2016) The default response to uncertainty and the importance of perceived safety in anxiety and stress: An evolution-theoretical perspective. J Anxiety Disord41, 22–34.

21.

Brosschot

, Verkuil

, Thayer

(2017) Exposed to events that never happen: Generalized unsafety, the default stress response, and prolonged autonomic activity. Neurosci Biobehav Rev74, 287–296.

22.

Amat

, Baratta

, Paul

, Bland

, Watkins

, Maier

(2005) Medial prefrontal cortex determines how stressor controllability affects behavior and dorsal raphe nucleus. Nat Neurosci8, 365–371.

23.

Maier

, Amat

, Baratta

, Paul

, Watkins

(2006) Behavioral control, the medial prefrontal cortex, and resilience. Dialogues Clin Neurosci8, 397–406.

24.

Motzkin

, Philippi

, Wolf

, Baskaya

, Koenigs

(2015) Ventromedial prefrontal cortex is critical for the regulation of amygdala activity in humans. Biol Psychiatry77, 276–284.

25.

Waldvogel

, van Gelderen

, Muellbacher

, Ziemann

, Immisch

, Hallett

(2000) The relative metabolic demand of inhibition and excitation. Nature406, 995–998.

26.

Jackson

(1884) The Croonian lectures on evolution and dissolution of the nervous system. Br Med J1, 660–663.

27.

Adam

, Kumari

(2009) Assessing salivary cortisol in large-scale, epidemiological research. Psychoneuroendocrinology34, 1423–1436.

28.

Gaffey

, Bergeman

, Clark

, Wirth

(2016) Aging and the HPA axis: Stress and resilience in older adults. Neurosci Biobehav Rev68, 928–945.

29.

Puig-Perez

, Almela

, Pulopulos

, Hidalgo

, Salvador

(2016) Are neuroticism and extraversion related to morning cortisol release in healthy older people?Int J Psychophysiol110, 243–248.

30.

Terracciano

, Stephan

, Luchetti

, Albanese

, Sutin

(2017) Personality traits and risk of cognitive impairment and dementia. J Psychiatr Res89, 22–27.

31.

Tautvydaite

, Kukreja

, Antonietti

, Henry

, von Gunten

, Popp

(2017) Interaction between personality traits and cerebrospinal fluid biomarkers of Alzheimer’s disease pathology modulates cognitive performance. Alzheimers Res Ther9, 6.

32.

Bennett

, McNeill

, Wolin

, Duncan

, Puleo

, Emmons

(2007) Safe to walk? Neighborhood safety and physical activity among public housing residents. PLoS Med4, 1599–1607.

33.

Lepore

, Kliewer

(2013) Violence exposure, sleep disturbance, and poor academic performance in middle school. J Abnorm Child Psychol41, 1179–1189.

34.

DeSantis

, Troxel

, Beckman

, Ghosh-Dastidar

, Hunter

, Hale

, Buysse

, Dubowitz

(2016) Is the association between neighborhood characteristics and sleep quality mediated by psychological distress? An analysis of perceived and objective measures of 2 Pittsburgh neighborhoods. Sleep Health2, 277–282.

35.

Duncan

, Park

, Goedel

, Kreski

, Morganstein

, Hambrick

, Jean-Louis

, Chaix

(2017) Perceived neighborhood safety is associated with poor sleep health among gay, bisexual, and other men who have sex with men in Paris, France. J Urban Health94, 399–407.

36.

Hek

, Direk

, Newson

, Hofman

, Hoogendijk

, Mulder

, Tiemeier

(2013) Anxiety disorders and salivary cortisol levels in older adults: A population-based study. Psychoneuroendocrinology38, 300–305.

37.

Johar

, Emeny

, Bidlingmaier

, Reincke

, Thorand

, Peters

, Heier

, Ladwig

(2014) Blunted diurnal cortisol pattern is associated with frailty: A cross-sectional study of 745 participants aged 65 to 90 years. J Clin Endocrinol Metab99, E464–E468.

38.

Noordam

, Jansen

, Akintola

, Oei

, Maier

, Pijl

, Slagboom

, Westendorp

, van der Grond

, de Craen

, van Heemst

(2012) Familial longevity is marked by lower diurnal salivary cortisol levels: The Leiden Longevity Study. PLoS One7, e31166.

39.

Fries

, Dettenborn

, Kirschbaum

(2009) The cortisol awakening response (CAR): Facts and future directions. Int J Psychophysiol72, 67–73.

40.

Chida

, Steptoe

(2009) Cortisol awakening response and psychosocial factors: A systematic review and meta-analysis. Biol Psychol80, 265–278.

41.

Bimonte-Nelson

, Acosta

, Talboom

(2010) Neuroscientists as cartographers: Mapping the crossroads of gonadal hormones, memory and age using animal models. Molecules15, 6050–6105.

42.

Belvederi Murri

, Pariante

, Mondelli

, Masotti

, Atti

, Mellacqua

, Antonioli

, Ghio

, Menchetti

, Zanetidou

, Innamorati

, Amore

(2014) HPA axis and aging in depression: Systematic review and meta-analysis. Psychoneuroendocrinology41, 46–62.

43.

Pineau

, Canet

, Desrumaux

, Hunt

, Chevallier

, Ollivier

, Belanoff

, Givalois

(2016) New selective glucocorticoid receptor modulators reverse amyloid-β peptide-induced hippocampus toxicity. Neurobiol Aging45, 109–122.

44.

Green

, Billings

, Roozendaal

, McGaugh

, LaFerla

(2006) Glucocorticoids increase amyloid-beta and tau pathology in a mouse model of Alzheimer’s disease. J Neurosci26, 9047–9056.

45.

Harris-White

, Chu

, Miller

, Simmons

, Teter

, Nash

, Cole

, Frautschy

(2001) Estrogen (E2) and glucocorticoid (Gc) effects on microglia and A beta clearance in vitro and in vivo. Neurochem Int39, 435–448.

46.

Sotiropoulos

, Catania

, Pinto

, Silva

, Pollerberg

, Takashima

, Sousa

, Almeida

(2011) Stress acts cumulatively to precipitate Alzheimer’s disease-like tau pathology and cognitive deficits. J Neurosci31, 7840–7847.

47.

Csernansky

, Dong

, Fagan

, Wang

, Xiong

, Holtzman

, Morris

(2006) Plasma cortisol and progression of dementia in subjects with Alzheimer-type dementia. Am J Psychiatry163, 2164–2169.

48.

Popp

, Wolfsgruber

, Heuser

, Peters

, Hüll

, Schröder

, Möller

, Lewczuk

, Schneider

, Jahn

, Luckhaus

, Perneczky

, Frölich

, Wagner

, Maier

, Wiltfang

, Kornhuber

, Jessen

(2015) Cerebrospinal fluid cortisol and clinical disease progression in MCI and dementia of Alzheimer’s type. Neurobiol Aging36, 601–607.

49.

Hixson

, Powers

(1991) Restriction isotyping of human apolipoprotein A-IV: Rapid typing of known isoforms and detection of a new isoform that deletes a conserved repeat. J Lipid Res32, 1529–1535.

50.

Fekedulegn

, Andrew

, Burchfiel

, Violanti

, Hartley

, Charles

, Miller

DBCM

(2007) Area under the curve and other summary indicators of repeated waking cortisol measurements. Psychosom Med69, 651–659.

51.

Pruessner

, Kirschbaum

, Meinlschmid

, Hellhammer

(2003) Two formulas for computation of the area under the curve represent measures of total hormone concentration versus time-dependent change. Psychoneuroendocrinology28, 916–931.

52.

Van Cauter

(1990) Diurnal and ultradian rhythms in human endocrine function: A minireview. Horm Res34, 45–53.

53.

Van Cauter

, Leproult

, Kupfer

(1996) Effects of gender and age on the levels and circadian rhythmicity of plasma cortisol. J Clin Endocrinol Metab81, 2468–2473.

54.

Stone

, Schwartz

, Smyth

, Kirschbaum

, Cohen

, Hellhammer

, Grossman

(2001) Individual differences in the diurnal cycle of salivary free cortisol: A replication of flattened cycles for some individuals. Psychoneuroendocrinology26, 295–306.

55.

Smyth

, Clow

, Thorn

, Hucklebridge

, Evans

(2013) Delays of 5–15min between awakening and the start of saliva sampling matter in assessment of the cortisol awakening response. Psychoneuroendocrinology38, 1476–1483.

56.

Smyth

, Thorn

, Hucklebridge

, Evans

, Clow

(2015) Detailed time course of the cortisol awakening response in healthy participants.. Psychoneuroendocrinology62, 200–203.

57.

Stalder

, Kirschbaum

, Kudielka

, Adam

, Pruessner

, Wüst

, Dockray

, Smyth

, Evans

, Hellhammer

, Miller

, Wetherell

, Lupien

, Clow

(2016) Assessment of the cortisol awakening response: Expert consensus guidelines. Psychoneuroendocrinology63, 414–432.

58.

Clow

, Thorn

, Evans

, Hucklebridge

(2004) The awakening cortisol response: Methodological issues and significanc. Stress7, 29–37.

59.

Charles

, Mogle

, Piazza

, Karlamangla

, Almeida

(2020) Going the distance: The diurnal range of cortisol and its association with cognitive and physiological functioning. Psychoneuroendocrinology112, 104516.

60.

van Coevorden

, Mockel

, Laurent

, Kerkhofs

, L’Hermite-Balériaux

, Decoster

, Nève

, Van Cauter

(1991) Neuroendocrine rhythms and sleep in aging men. Am J Physiol260, E651–661.

61.

Evans

, Fredhoi

, Loveday

, Hucklebridge

, Aitchison

, Forte

, Clow

(2011) The diurnal cortisol cycle and cognitive performance in the healthy old. Int J Psychophysiol79, 371–377.

62.

Heaney

, Phillips

, Carroll

(2010) Ageing, depression, anxiety, social support and the diurnal rhythm and awakening response of salivary cortisol. Int J Psychophysiol78, 201–208.

63.

Dmitrieva

, Almeida

, Dmitrieva

, Loken

, Pieper

(2013) A day-centered approach to modeling cortisol: Diurnal cortisol profiles and their associations among U.S. adults. Psychoneuroendocrinology38, 2354–2365.

64.

Lederbogen

, Kühner

, Kirschbaum

, Meisinger

, Lammich

, Holle

, Krumm

, von Lengerke

, Wichmann

, Deuschle

, Ladwig

(2010) Salivary cortisol in a middle-aged community sample: Results from 990 men and women of the KORA-F3 Augsburg study. Eur J Endocrinol163, 443–451.

65.

Miller

, Chen

, Zhou

(2007) If it goes up, must it come down? Chronic stress and the hypothalamic-pituitary-adrenocortical axis in humans. Psychol Bull133, 25–45.

66.

Wrosch

, Miller

, Lupien

, Pruessner

(2008) Diurnal cortisol secretion and 2-year changes in older adults’ physical symptoms: The moderating roles of negative affect and sleep. Health Psychol27, 685–693.

67.

Kumari

, Shipley

, Stafford

, Kivimaki

(2011) Association of diurnal patterns in salivary cortisol with all-cause and cardiovascular mortality: Findings from the Whitehall II study. J Clin Endocrinol Metab96, 1478–1485.

68.

Brosschot

(2017) Ever at the ready for events that never happen. Eur J Psychotraumatol8, 1309934.

69.

Maren

, Phan

, Liberzon

(2013) The contextual brain: Imlications for fear conditioning, extinction and psychopathology. Nat Rev Neurosci14, 417–428 .

70.

Pappens

, Schroijen

, Sütterlin

, Smets

, Van den Bergh

, Thayer

, Van Diest

(2014) Resting heart rate variability predicts safety learning and fear extinction in an interoceptive fear conditioning paradigm. PLoS One9, e105054.

71.

Quirk

, Gehlert

(2003) Inhibition of the amygdala: Key to pathological states?Ann N Y Acad Sci985, 263–272.

72.

Smith

, Allen

JJB

, Thayer

, Lane

(2015) Altered functional connectivity between medial prefrontal cortex and the inferior brainstem in major depression during appraisal of subjective emotional responses: A preliminary study. Biol Psychol108, 13–24.

73.

Van Marle

, Hermans

, Qin

, Fernández

(2010) Enhanced resting-state connectivity of amygdala in the immediate aftermath of acute psychological stress. Neuroimage53, 348–354.

74.

Grupe

, Nitschke

(2013) Uncertainty and anticipation in anxiety: An integrated neurobiological and psychological perspective. Nat Rev Neurosci14, 488–501.

75.

Clancy

, Prestwich

, Caperon

, O’Connor

(2016) Perseverative cognition and health behaviors: A systematic review and meta-analysis. Front Hum Neurosci10, 534.

76.

Goodman

, Bruce

, Cheng

, Mattson

(1996) Estrogens attenuate and corticosterone exacerbates excitotoxicity, oxidative injury, and amyloid beta-peptide toxicity in hippocampal neurons. J Neurochem66, 1836–1844.

77.

Popp

, Schaper

, Kölsch

, Cvetanovska

, Rommel

, Klingmüller

, Dodel

, Wüllner

, Jessen

(2009) CSF cortisol in Alzheimer’s disease and mild cognitive impairment. Neurobiol Aging30, 498–500.

78.

Lee

, Glass

, McAtee

, Wand

, Bandeen-Roche

, Bolla

, Schwartz

(2007) Associations of salivary cortisol with cognitive function in the Baltimore memory study. Arch Gen Psychiatry64, 810–818.

79.

Dos Santos

, Leyendecker

DMD

, Costa

ALS

, de Souza-Talarico

(2018) Relationship between cortisol reactivity to psychosocial stress and declarative memory decline during aging: Impact of age and sex. Geriatr Gerontol Int18, 169–176.