The biology begins with the hypothalamic-pituitary-adrenal axis, the body’s central stress machinery. When the brain detects a threat, real or perceived, the hypothalamus triggers a hormonal cascade that ends with the adrenal glands flooding the bloodstream with glucocorticoids—cortisol in humans, corticosterone in rodents. These hormones mobilize energy for survival, sharpen performance, and then shut themselves down through negative feedback, binding to glucocorticoid and mineralocorticoid receptors in the brain to suppress further release. When this feedback system is disrupted by intense, repeated, or prolonged stress, the result can be the dysregulated arousal, immune function, and emotional processing seen in stress-related disorders such as PTSD.

What has long puzzled scientists is the selectivity of the damage. Many individuals experience trauma; only a fraction develop pathology. Human studies have correlated pre- and post-stress cortisol levels with PTSD development and symptom severity, and rodent work has linked corticosterone to PTSD-like behavioral deficits—but most of those experiments were conducted in males only, and it remained unknown whether stress reactivity itself could forecast vulnerability before trauma strikes. The Cincinnati team, led by Emily Devine and James Herman, set out to answer exactly that question, with sex as a central variable rather than an afterthought.

The experimental design was deliberately longitudinal. Young adult male and female Sprague-Dawley rats first provided baseline morning and afternoon corticosterone measurements via tail blood samples. Three days later, each animal was placed in a light-dark box—a mild anxiogenic challenge in which rodents explore an illuminated open compartment and a dark enclosed one—and blood was drawn at 15, 30, 60, and 120 minutes afterward. The hormone values were integrated into an area-under-the-curve score capturing the total magnitude of each animal’s response. Crucially, behavioral analysis of the light-dark test itself showed no group differences in time spent in either compartment, latency to cross, or velocity, meaning the hormone signal reflected genuine physiological responsivity rather than differences in how the animals behaved during the task.

Three weeks later, the

Three weeks later, the animals assigned to the traumatic stress condition underwent single-prolonged stress, a widely used rodent model that combines restraint, forced swim, and ether exposure in a single session, followed by a period of quiet recovery. The control animals remained undisturbed in their home cages. This three-week interval matters scientifically: it allowed any predictive hormone signature measured on the mild-stress day to fade from acute influence, so that associations emerging later could plausibly reflect stable individual traits rather than lingering effects of the earlier test. After the traumatic stress exposure, the researchers tracked morning corticosterone at one and seven days, then ran a battery of behavioral assays ordered from least to most stressful, with rest days in between to minimize carryover effects.

The behavioral results painted a coherent picture in females. Animals exposed to single-prolonged stress showed reductions in basal morning corticosterone, echoing a pattern often reported in humans with PTSD, where basal cortisol tends to sit below that of healthy controls even as stress-system feedback is enhanced. Among the female traumatic stress animals, those classified as high corticosterone responders on the mild-stress day displayed more anxiety-like behavior, poorer fear extinction, and stronger reinstatement of fear—essentially a full traumatic stress-related phenotype. In plain terms, the females whose hormonal systems had reacted most vigorously to a trivial challenge weeks earlier were the same ones that struggled most to let go of fear after trauma.

The statistical approach strengthened this claim. Beyond dividing animals into upper and lower quartile responders, the team ran correlational and permutation analyses across the full cohort of twenty animals per condition. In females exposed to traumatic stress, a consistent pattern emerged: lower pre-stress basal morning corticosterone combined with a larger hormone response to the mild stressor predicted greater fear expression. This dissociation between resting and stress-evoked glucocorticoid signals is conceptually important. A low resting level with a high reactive surge suggests an HPA axis that is quiet at baseline but quick to overshoot when provoked—a profile that appears to confer vulnerability in females rather than protection.

Males told a different story. Mild stress reactivity did not predict any of the traumatic stress-related behavioral outcomes in males, suggesting that the size of an acute glucocorticoid response carries little prognostic information for them under these conditions. Yet the data were not silent: in control males—animals never exposed to the traumatic protocol—reduced pre-stress basal morning corticosterone predicted increased fear expression during testing. The predictive value of basal hormone levels thus flipped context depending on sex and experience, hinting that the same physiological measure can index different underlying processes in different biological and environmental settings.

The mechanistic layer of the study came from cFos immunohistochemistry, a technique that reveals which neurons were recently active by detecting expression of an immediate-early gene. Brains were collected ninety minutes after fear reinstatement, allowing the researchers to map neural engagement at the moment fearful memories were being reactivated. Traumatic stress exposure and corticosterone responder status each shaped cFos expression across a network of stress-relevant regions: the paraventricular nucleus of the hypothalamus, the hub that drives HPA axis output; the infralimbic cortex, a medial prefrontal region essential for extinguishing conditioned fear in rodents; and subregions of the amygdala, the threat-detection structure whose hyperactivity is a hallmark of PTSD. Differential activation patterns in these areas provide a plausible neural bridge between hormone dynamics measured weeks earlier and behavior measured later.

The choice of a light-dark box as the mild stressor deserves emphasis. Unlike foot shocks, restraint, or predator odor, the light-dark box creates psychological novelty and mild anxiety without physical harm, and the researchers confirmed behaviorally that the animals’ exploration did not differ systematically across groups. This makes the corticosterone measurements a relatively clean readout of internal physiological reactivity—how strongly an individual’s stress axis responds to a given situation—rather than a confound of gross behavioral differences. Quantifying the response as area under the curve across four time points also captured the full shape of the hormone trajectory, not merely a single peak value, which is a more faithful description of glucocorticoid dynamics as endocrinologists understand them.

The sex-specific findings sit within a broader scientific conversation. Women are diagnosed with PTSD at roughly twice the rate of men, a disparity long attributed to some mixture of trauma exposure differences, social factors, and biology. Rodent studies, which can control environment and trauma exposure precisely, offer a way to isolate the biological component. Much of the earlier rodent literature on glucocorticoids and trauma vulnerability relied on male animals, a gap this study explicitly addressed by running matched male and female cohorts with identical timelines. The fact that a predictive relationship emerged robustly in females while absent in males—under carefully controlled conditions where both sexes experienced identical procedures—suggests that glucocorticoid responsivity may be a sex-dependent biomarker rather than a universal one.

The interpretation of low basal corticosterone also benefits from context. In the PTSD literature, low cortisol has been proposed not simply as a consequence of the disorder but as a potential pre-existing risk factor, an idea supported by studies of trauma survivors assessed before or shortly after their trauma. The present findings partially echo this: in females, lower pre-stress basal levels combined with higher reactivity predicted worse outcomes, while in untraumatized males low basal levels likewise predicted more fear expression. What complicates any simple reading is that the same basal measure related to different outcomes depending on sex and trauma history, underscoring that glucocorticoid biology is not linear—both insufficient and excessive signaling can be maladaptive, and the meaning of any single number depends on the surrounding regulatory context.

The three-week delay between mild stress testing and traumatic stress also raises interesting questions about trait stability. Individual differences in HPA axis reactivity are thought to reflect some combination of genetics, early life experience, and epigenetic tuning of glucocorticoid receptor expression in the hippocampus and hypothalamus. That a single mild challenge predicted behavior three weeks later implies a reasonably stable individual characteristic, at least across this window of young adulthood in the rat. Whether the same stability holds across longer timespans, aging, or additional stress exposure remains open.

Several caveats temper the translational enthusiasm. The quartile-based responder analysis involved small groups of five animals each, even though the full-cohort correlations provided convergent support. Rats are not humans, corticosterone is not cortisol, and the single-prolonged stress model, while valuable, captures only selected features of human trauma. Neither can the study resolve whether glucocorticoid dynamics cause vulnerability or merely correlate with some upstream factor—perhaps differences in prefrontal or amygdala circuitry—that drives both hormone regulation and fear behavior. Still, the central achievement stands: a simple, non-invasive, pharmacology-free measurement taken weeks before trauma predicted downstream outcomes in females, with converging neural activation data. If similar predictive relationships hold in humans, a morning cortisol assessment paired with a standardized mild stressor could one day help identify which individuals deserve heightened monitoring or early intervention after traumatic events, before symptoms ever appear.

Subject of Research: Sex-specific prediction of traumatic stress susceptibility by glucocorticoid dynamics in rats

Article Title: Glucocorticoid dynamics predict susceptibility and resilience to traumatic stress in male and female rats

Article References: Devine, E. A., Chambers, J. B., Dunn, T., Musarra, A., Porter, S., Moore, C. E., Smail, M., Nawreen, N., Packard, B. A., & Herman, J. P. (2026). Glucocorticoid dynamics predict susceptibility and resilience to traumatic stress in male and female rats. Frontiers in Cellular Neuroscience, 20, Article 1880559. https://doi.org/10.3389/fncel.2026.1880559

Image Credits: AI Generated

DOI: 10.3389/fncel.2026.1880559

Keywords: PTSD, corticosterone, HPA axis, single prolonged stress, fear extinction, sex differences, biomarker, glucocorticoids, amygdala, hypothalamus, rat model, stress reactivity