Stress System Concordance: A Signature of Resilience

Stress causes depression in some individuals, but others are resilient. This proposition has abundant supporting evidence, but predicting good and bad stress outcomes is still cumbersome. In Biological Psychiatry: Global Open Science, Kuhn et al. (1) report that concordance in various readouts of the psychophysiological stress response is a signature of resilience, and discordance is a feature of depression. But what is stress concordance, and when does it become discordant?

SEE CORRESPONDING ARTICLE NO. 100400

Kuhn et al. (1) operationalized concordance based on the within-person clustering of 3 variables: 1) stress-induced cortisol (CORT) response, 2) self-reported, and 3) experimenter-observed experiences of positive and negative affect. Categorization was not theory based but data driven and derived from objective assessment. The number of concordance categories was a priori set at a maximum of 4 because of statistical power. The authors also aimed to substantiate the construct validity of the concordance categorization by predicting categorical membership with relevant variables such as exposure to early-life trauma and previous or current depression. The strength of this innovative approach lies in its ability to assess the coherence of reactivity to stress across several domains, leading to findings that a less sophisticated approach would not have picked up.

The critical element in the study is the use of the Maastricht Acute Stressor Test, a test similar to the Trier Social Stress Test that measures the stress response to a socioevaluative stressor (2). To this end, the investigators collected saliva samples to measure free CORT, which is not bound to corticosteroid-binding globulin as in blood. Unfortunately, the study sample was not large enough to test for sex differences in CORT reactivity, which is known to exist from meta-analytic evidence (3). The subsequent Maastricht Acute Stressor Test triggered a more prolonged and attenuated CORT response than was observed in the original test, where CORT peaks twice as high from baseline at 20 minutes. The reason for this apparent discrepancy is obvious: the authors measured the CORT response during the experience of the brain image analysis procedure. In the future, the reliability of baseline and reactivity measurements will significantly improve with the application of real-time ambulant continuous CORT sampling by microdialysis in subcutaneous fat recently developed by Stafford Lightman and collaborators.

The study raises several interesting perspectives. The first issue regards predisposition for the physiology of stress system dis- and concordance. Pham et al. (4) addressed this in previous research by showing that an asymmetrical coactivation pattern of CORT and α-amylase, a marker of sympathetic nervous system activity, is associated with higher exposure to chronic stress and poorer mental health outcomes during adolescence. Such asymmetry as a manifestation of discordance may be of genetic origin because extreme differences in coping styles coexist: sympathetic nervous system–driven fight-flight versus CORT-dependent conservation withdrawal. In studies by Jaap Koolhaas and colleagues with genetically selected rodents, the phenotype with a high sympathetic nervous system/low corticosterone ratio (short attack latency mice) dominates when in control in the home environment but not when exposed to novelty. Therefore, contextual aspects are another factor that needs attention in the concordance concept.

That individuals with a discordant hyper- or hyporeactive CORT response pattern displayed significant associations with depression severity, anhedonia, and childhood adversity is an additional outcome from Kuhn’s study (1). Also, in animal studies, early-life stress programs the stress system for life. For example, days-old rat pups become stress responsive upon maternal separation. Such a stressful experience during the early postnatal stress hyporesponsive period produces an untimely rise in corticosterone secretion, which stimulates the growth of the amygdala and enhances emotional reactivity later in life. This finding has been confirmed with early-life experience in humans and led to the predictive adaptive response or 3-hit hypothesis of resilience and vulnerability. Depending on genetic background, early-life stress inoculation promotes resilience to similar later-life stressors. Resiliency is compromised when a mismatch (discordance) between this early-life epigenetic signature and the context of later-life experience occurs (5).

The second issue involves the cause of changes in CORT secretion. CORT, the end-product of the hypothalamic-pituitary-adrenal axis, acts in concert with its various neuropeptides (corticotropin-releasing hormone, opiomelanocortins, oxytocin, and vasopressin) to promote adaptation. The CORT hyporeactivity and high affective reactivity observed by Kuhn et al. (1) may be due to primarily diminished adrenal CORT secretion or a consequence of decreased processing of the internal affective state. If it is the former, there may be changes at any point in the hypothalamic-pituitary-adrenal axis to explain hypocortisol. These include the variation in adrenal sensitivity to adrenocorticotropic hormone and diminished mitochondrial CORT synthesis because of a lack of energy substrates, as pointed out by Clemens Kirschbaum and colleagues. At the pituitary adrenocorticotropic hormone release level, the dexamethasone suppression test can identify variations in the negative feedback of CORT in the face of a hypothalamic corticotropin-releasing hormone/vasopressin drive.

CORT also feeds back on the brain’s social decision network underlying the selection of coping style. CORT targets this circuitry after activating its mineralocorticoid- and glucocorticoid receptors (GRs) via complementary genomic and nongenomic mechanisms that involve regulatory factors such as the FKBP5 cochaperone and the tissue-specific coregulators. Mineralocorticoid receptor–mediated actions may modulate memory retrieval and coping style. GR-mediated actions regulate perception, appraisal processes, (immune) defense reactions, memory consolidation, and energy metabolism, all serving adaptation. Therefore, these receptors regulate different phases of the stress-induced CORT response: Mineralocorticoid receptor activation is concerned with the initial phase, while GR determines its duration. In other words, rapid activation of the CORT response promotes resilience as long as its duration is also efficiently terminated (6).

The third issue involves the course of CORT action underlying stress coping and adaptation. In Kuhn’s study (1), CORT hyperreactivity in vulnerable individuals and hyporeactivity as a characteristic feature of the individual with depression follows the pattern previously described for the course of depression severity over time (Figure 1). The initial phase of CORT hyperactivity and high affective reactivity, termed “acute stress syndrome,” switches to what Robert Dantzer coined “sickness syndrome,” which is characterized by CORT hyporeactivity and lack of energy. Defunct mitochondria and a proinflammatory phenotype are at the root of this downward spiral of energy depletion (7). Alternatively, a recent study of rodents suggests the existence of a GR-dependent switch in ascending serotonergic neurons from excitatory glutamatergic to inhibitory GABAergic (gamma-aminobutyric acidergic) transmission upon chronic stress exposure, a finding that has been confirmed in the postmortem human brain (8). The chain of events during chronic stress exposure toward depression is fundamental to Bruce McEwen’s allostatic load model. It is also reminiscent of Selye’s general adaptation syndrome, where a resistance phase precedes exhaustion.

Figure 1.

Illustration of the cortisol (CORT)/affective response trajectories presented by Kuhn et al. (1). (A) The normative profile, with low CORT levels in saliva and low negative affect (blue lines) and high positive affect (green lines) at baseline and a moderate increase in CORT levels and stable affect after stress exposure. (B) The CORT hyperreactivity/affect stability profile, with high CORT levels in saliva and high negative affect and low positive affect at baseline and a significant increase in CORT levels and stable affect after stress exposure. (C) The CORT hyporeactivity/affect reactivity profile, with high CORT levels in saliva and high negative affect and low positive affect at baseline and a hyporeactive CORT response and increases in negative mood and decreases in positive mood stable affect after stress exposure. Note that Kuhn et al. (1) reported 4 different profiles. The 2 CORT hyporeactivity/affect reactivity profiles are presented here as a single category due to apparent similarities between phenotypes and for reasons of simplicity. The timeline represents the onset and progression of depression. In the early phases, CORT and affect are reactive to stress (the normative profile). Prolonged CORT hyperactivity can result in acute stress syndrome (allostatic overload). Over time, a switch occurs into the sickness syndrome, which is characterized by low energy levels, a proinflammatory phenotype, anhedonia, and a hyporeactive CORT response (7).

Then, with the current state of the art, what would be the spinoff from a CORT-centered perspective for diagnosing and treating patients suffering from stress-related affective disorders? With this question and current findings on dis-/concordance in the psychophysiology of CORT action, Hellhammer’s hypothetical conceptual endophenotypes and Neuropattern come to mind (9). Conceptual endophenotypes based on patterns of psychophysiological and symptomatic elements extending over different levels of biological complexity may serve to identify subgroups of patients with affective disorders. Corticotropin-releasing hormone hypoactivity, for example, is characterized by a low CORT level and symptoms of affective reactivity and sickness syndrome with a proinflammatory cytokine profile, quite similar to the phenotype noted by Kuhn et al. (1).

What next? Firstly, the imaging part of the current study will reveal how the imbalance in physiology and perception of stress in the activity of brain circuits that underlie cognitive control over appraisal processes and emotional reactivity is related to central CORT action. Secondly, regarding research on molecular underpinning, the very recent giant integrated multi-omics analysis from protein to DNA of postmortem medial prefrontal cortex, central amygdala, and hippocampal dentate gyrus points to interleukin 1β, GR, STAT3, and tumor necrosis factor as top-stream associates of depression and posttraumatic stress disorder (10) supporting additional research on affective disorders following a CORT-centered approach that dissects the gene × environment interactions at play. Thirdly, participation in worldwide data sharing and collaborative initiatives, including neurobiological, physiological, and behavioral data on stress vulnerability and resilience, is encouraged. See STRESS-NL database (https://www.stressdatabase.eu) and the Global Stress and Resilience Network (https://www.stressnetwork.ch).