Cortisol And Depression Clinical Research

A comprehensive review for psychiatry researchers, clinicians, and informed readers

---

Table of Contents

• What Is the HPA Axis and Why Does It Matter for Depression?
• Cortisol and Depression: The Core Clinical Question
• What the Latest Research Says (2024–2026)
• Morning vs. Nighttime Cortisol in Depression
• Is High Cortisol a Cause or Consequence of Depression?
• Low Cortisol and Depression: The Blunted Hypothesis
• Cortisol in Adolescents and Young Adults With Depression
• Cortisol, Suicidal Behavior, and Severe Depression
• Can Cortisol Tests Diagnose Depression?
• Cortisol as a Predictor of Relapse and Treatment Response
• Antidepressants and Cortisol: What Happens to HPA Activity?
• Cortisol and Anxiety vs. Major Depression: An Important Distinction
• Clinical Implications and the Road Ahead
• Frequently Asked Questions

---

Introduction

For decades, psychiatry researchers have pursued a deceptively simple question: does cortisol — the body's primary stress hormone — hold the key to understanding major depressive disorder? The short answer is nuanced, deeply contested, and more clinically important today than ever before.

Cortisol and depression clinical research has produced a large, complex, and sometimes contradictory body of evidence. Some studies show that depressed patients have chronically elevated cortisol. Others suggest that in certain forms of depression, cortisol is actually lower than normal. A landmark 2025 Mendelian randomization study added another layer of complexity by suggesting depression itself may cause blunted morning cortisol — reversing the assumed direction of causality that had guided research for decades.

This post synthesizes the current state of evidence across multiple dimensions: mechanistic research on the HPA axis, clinical biomarker data, cutting-edge 2024–2026 publications, and the practical implications for psychiatrists, psychologists, and translational researchers. Whether you are evaluating a patient, designing a study, or reviewing grant literature, this overview is designed to serve as a rigorous, evidence-based reference.

---

---

What Is the HPA Axis and Why Does It Matter for Depression?

The hypothalamic-pituitary-adrenal (HPA) axis is the central neuroendocrine system governing the body's response to psychological and physiological stress. Understanding HPA axis depression connections begins with knowing how this system actually works at a mechanistic level.

The Basic Circuit

1. Hypothalamus detects stress signals and releases corticotropin-releasing hormone (CRH) and arginine vasopressin (AVP).
2. Anterior pituitary responds to CRH/AVP by secreting adrenocorticotropic hormone (ACTH) into the bloodstream.
3. Adrenal cortex receives ACTH and releases cortisol into systemic circulation.
4. Negative feedback loop: Cortisol then travels back to glucocorticoid receptors (GRs) in the hippocampus, prefrontal cortex, and hypothalamus, signaling the system to shut down.

In healthy individuals, this circuit is tightly regulated, producing a predictable diurnal pattern: cortisol peaks within 30–45 minutes of waking (the cortisol awakening response, or CAR), declines across the morning, drops sharply in the afternoon, and reaches a nadir in the early hours of sleep.

Why HPA Dysfunction Is Central to Psychiatric Illness

The HPA depression relationship has been a cornerstone of biological psychiatry since at least the 1960s, when researchers first noted that patients with endogenous depression showed elevated urinary cortisol and failed to suppress cortisol normally in response to the synthetic glucocorticoid dexamethasone — the so-called Dexamethasone Suppression Test (DST).

The logic is intuitive: if the brain's stress response system is dysregulated, it would logically affect neurotransmitter systems, neuroplasticity, immune function, sleep architecture, and appetite — precisely the domains disrupted in major depressive disorder (MDD). Glucocorticoid receptors are densely expressed in limbic and prefrontal regions that regulate emotion, threat appraisal, and executive function. Chronic HPA hyperactivation can:

• Suppress hippocampal neurogenesis and reduce hippocampal volume
• Impair prefrontal cortex function and cognitive flexibility
• Alter serotonergic, dopaminergic, and noradrenergic neurotransmission
• Drive systemic inflammation through complex glucocorticoid-immune interactions
• Disrupt circadian rhythms by uncoupling the cortisol diurnal cycle from the central clock

Understanding the cortisol depression mechanism therefore requires appreciating that cortisol is not simply a downstream stress hormone — it is an active modulator of the very neural systems that depression affects most profoundly.

Key Terminology for Researchers

| Term | Definition | |------|-----------| | CAR (Cortisol Awakening Response) | Cortisol spike 0–45 min post-waking; index of HPA reactivity | | DST (Dexamethasone Suppression Test) | Tests negative feedback integrity; non-suppression = HPA dysregulation | | ACTH | Pituitary hormone stimulating adrenal cortisol release | | CRH | Hypothalamic hormone driving ACTH secretion | | GR (Glucocorticoid Receptor) | Nuclear receptor mediating cortisol's feedback and cellular effects | | HPA hyperreactivity | Exaggerated cortisol response to stress or ACTH stimulation | | HPA hypoactivity / blunted cortisol | Attenuated cortisol production, seen in some depression subtypes |

---

Cortisol and Depression: The Core Clinical Question

The foundational question driving cortisol depression research is straightforward to state and extraordinarily difficult to resolve: Are cortisol levels reliably abnormal in major depression, and if so, in what direction?

Decades of research have produced three broad findings that, taken together, seem contradictory until you account for methodological and clinical heterogeneity:

Finding 1: Many patients with MDD show elevated basal and stress-reactive cortisol. This is the classic "hypercortisolism" finding, most robustly documented in melancholic, severe, or psychotic depression and in older adult samples.

Finding 2: Some patients with depression, particularly those with atypical depression, chronic fatigue comorbidity, or PTSD comorbidity, show hypocortisolism — blunted or flat cortisol profiles.

Finding 3: A 2025 Mendelian randomization analysis suggests depression itself causes lower morning plasma cortisol (β = −0.107, 95% CI −0.181 to −0.032), meaning what looks like a predisposing factor may actually be a consequence of the depressive state.

These findings are not necessarily mutually exclusive. They likely reflect the biological heterogeneity of what clinicians call "depression" — a syndromal diagnosis that almost certainly encompasses multiple pathophysiologically distinct entities with divergent HPA phenotypes.

Why the Research Is So Hard to Synthesize

Cortisol MDD research faces several methodological challenges that make simple conclusions dangerous:

• Sampling timing matters enormously. Cortisol levels vary by an order of magnitude across the diurnal cycle. A morning sample, an afternoon sample, and an evening sample are measuring fundamentally different things.
• Sampling medium varies. Plasma, serum, urine (24-hour free cortisol), saliva, and hair (reflecting weeks to months of cumulative secretion) are not equivalent measures.
• Depression subtypes are rarely stratified. Melancholic, atypical, psychotic, and treatment-resistant depression have distinct HPA profiles that pooling obscures.
• Confounders are numerous. Body mass index, alcohol use, smoking, sleep quality, physical activity, age, sex, antidepressant use, and trauma history all independently affect cortisol.
• Publication bias tends to favor positive findings.

With that methodological context established, let us turn to what the most rigorous and recent evidence actually shows.

---

What the Latest Research Says (2024–2026)

The past two years have produced some of the most methodologically sophisticated work in the cortisol mood disorder research field to date. Here is a synthesis of the key 2024–2026 publications.

The 2024 Meta-Analysis on Adolescents and Young Adults

A 2024 meta-analysis and systematic review focusing on depressed adolescents and young adults found that this population showed significantly higher cortisol levels than controls, with a standardized mean difference (SMD) of 0.87 (95% CI 0.43–1.31). This is a medium-to-large effect size by conventional standards and one of the larger effect sizes seen in the adolescent literature.

Critically, the same analysis found a higher cortisol stress response in depressed youth, with SMD 0.68 (95% CI 0.31–1.05). This suggests that the elevated cortisol is not merely a trait phenomenon but is also manifest in exaggerated reactivity to acute stressors — consistent with HPA sensitization models.

Perhaps most strikingly, this 2024 meta-analysis reported a positive linear association between morning/afternoon cortisol and depression symptom scores in adolescents and young adults, with a correlation coefficient of r = 0.82. This is a remarkably strong association for a biological marker in psychiatry, though it should be interpreted cautiously given the potential for confounding and the range of methodological approaches across included studies.

Clinical implication: The cortisol-depression relationship may be particularly strong and measurable in younger populations, possibly because confounders (chronic illness, medication burden, HPA senescence) are less prominent. This argues for targeted depression cortisol study designs in adolescent cohorts.

The 2024 Pooled Analysis: Anxiety vs. Major Depression

A 2024 pooled analysis yielded a finding that has generated considerable discussion in cortisol psychiatric research circles: cortisol was significantly associated with anxiety (OR 1.16 per cortisol z-score, 95% CI 1.04–1.31), but was not significantly associated with major depression (OR 1.02, 95% CI 0.95–1.10).

This is a pivotal finding for several reasons:

1. It challenges the assumption that elevated cortisol is a reliable general feature of MDD in population-level analyses.
2. It raises the possibility that elevated cortisol may be more etiologically relevant to anxiety disorders — which frequently co-occur with MDD — than to depression per se.
3. It is consistent with the view that MDD is biologically heterogeneous and that cortisol elevation may characterize a specific subtype rather than the syndrome as a whole.

Researchers should note, however, that such pooled analyses may mask subtype-specific effects and are limited by how "major depression" is operationalized across contributing studies.

The 2025 Mendelian Randomization Analysis

Mendelian randomization (MR) uses genetic variants as instrumental variables to probe causal relationships, bypassing many of the confounding problems that plague observational research. The 2025 MR analysis in this space found that depression was inversely associated with morning plasma cortisol (β = −0.107, 95% CI −0.181 to −0.032).

This finding is important for two interconnected reasons:

First, it supports the hypothesis that depression causes blunted cortisol — not the reverse. This is a significant challenge to traditional HPA hypercortisolism models that assume elevated cortisol contributes to depression onset.

Second, it aligns with the blunted HPA reactivity findings documented in atypical depression, post-traumatic states, and chronic depressive illness, suggesting these may represent downstream HPA exhaustion or allostatic adaptation rather than independent predisposing pathology.

As with all MR studies, interpretation requires caution: the genetic variants used as instruments for depression risk may have pleiotropic effects, and the analysis may not capture the full heterogeneity of depression subtypes.

The 2025 Findings on Suicidal Behavior

A 2025 meta-analysis and case-control study examining depression with suicidal behavior found that cortisol levels in this population were higher than in healthy controls. This adds specificity to the cortisol-depression picture: while cortisol may not differ from controls in all forms of MDD, the subpopulation presenting with active suicidal ideation or behavior may represent a high-cortisol phenotype.

This has potential clinical significance as a severity or risk marker, though the research is not yet at a stage where it can inform routine clinical decision-making about suicide risk.

The 2025 Review on Circadian Dysregulation and Mood Disorders

A 2025 updated review described cortisol dysregulation as linked to mood disorders and depression-related circadian and feedback abnormalities, reinforcing the view that it is not just the level of cortisol that matters in depression, but the patterning — the diurnal profile, the feedback sensitivity, and the alignment of cortisol rhythms with sleep-wake cycles.

This framing shifts the research question from "how much cortisol?" to "what is the shape of the cortisol curve, and how does it relate to circadian disruption?" — a more nuanced and likely more clinically meaningful formulation.

---

---

Morning vs. Nighttime Cortisol in Depression

One of the most important — and most frequently overlooked — distinctions in HPA psychiatric research is the difference between morning cortisol and nighttime/evening cortisol in depression. These are not simply different time points on the same curve; they reflect different regulatory mechanisms and have different clinical implications.

Morning Cortisol and the Cortisol Awakening Response (CAR)

The cortisol awakening response (CAR) — the rapid surge in cortisol that occurs in the 30–45 minutes after waking — is thought to reflect HPA axis anticipatory reactivity and hippocampal-HPA feedback sensitivity. It is regulated partly by the central circadian clock and partly by psychological appraisal of upcoming demands.

In melancholic and psychotic depression: The CAR is often elevated, consistent with chronic HPA hyperactivation. Patients may wake with high cortisol and sustain elevated levels through the morning.

In atypical depression and burnout-related depression: The CAR may be flattened or blunted, consistent with HPA hypofunction. These patients often report profound fatigue, hypersomnia, and hyperphagia — a clinical phenotype that fits low-arousal, rather than high-arousal, HPA activity.

The 2025 Mendelian randomization finding of lower morning plasma cortisol in depression (β = −0.107) may specifically be capturing this blunted-CAR phenotype at a population level, which could explain why depression's genetic liability predicts lower, not higher, morning cortisol in an unselected sample.

Nighttime Cortisol: The Disinhibition Signature

In healthy individuals, nighttime cortisol is near-undetectable. The hypothalamus effectively suppresses ACTH secretion during slow-wave sleep, allowing cortisol to reach its nadir. In severe depression, particularly melancholic depression, nighttime cortisol may be elevated — reflecting failure of the normal nocturnal HPA suppression.

This nocturnal cortisol elevation can:

• Fragment sleep architecture, reducing slow-wave sleep
• Accelerate hippocampal glucocorticoid exposure, potentially contributing to volume loss
• Dysregulate immune cytokine production, which follows a nocturnal pattern

Elevated late-night or midnight cortisol is one of the more robust biological signatures of severe, endogenous MDD and was historically captured by the DST (failure to suppress overnight cortisol in response to dexamethasone given at 11 PM).

Clinical Research Implication

For cortisol depression research to be interpretable, sampling protocols must specify:

1. Whether morning CAR, daytime, afternoon, or nighttime cortisol is being measured
2. The sampling medium (plasma, saliva, urine, hair)
3. The timing relative to waking, meals, and activity
4. Whether the CAR area under the curve (AUC) is being computed for reactivity vs. total output

Studies that conflate these measures will produce heterogeneous results that are difficult to synthesize — which is precisely what has plagued this literature for decades.

---

Is High Cortisol a Cause or Consequence of Depression?

This is arguably the central unresolved question in the cortisol depression mechanism literature: does HPA hyperactivation cause depression, or does depression cause HPA dysregulation? The answer matters enormously for therapeutic strategy and biomarker development.

The Causation Hypothesis (Cortisol → Depression)

The classical model, dominant from the 1970s through the 2000s, held that chronic stress → HPA hyperactivation → elevated glucocorticoid exposure → hippocampal damage, monoamine depletion, and eventually clinical depression. Key supporting evidence:

• Cushing's syndrome (pathological cortisol excess from adrenal tumor or exogenous steroids) produces depression in a large minority of patients.
• Exogenous corticosteroid treatment is associated with mood disturbances ranging from euphoria to severe depression.
• Animal models of chronic stress produce HPA hyperactivation alongside depression-like behavioral phenotypes that can be rescued by glucocorticoid receptor antagonism.
• Early-life adversity (childhood trauma, maltreatment) programs HPA hyperreactivity that persists into adulthood and predicts depression onset.

The Consequence Hypothesis (Depression → Cortisol Dysregulation)

The 2025 Mendelian randomization evidence points in the opposite direction for morning cortisol specifically. A 2019 systematic review and meta-analysis also found that cortisol predicted MDD onset/relapse/recurrence with OR 1.294 (95% CI 1.035–1.616), but the authors themselves noted significant confounding by baseline depression severity and study quality — meaning even this "predictive" finding may partly reflect pre-existing or subsyndromal depression affecting the cortisol measurement.

Under this framing, cortisol dysregulation is not a driver of depression but a biological signature or consequence of the depressive state — a reflection of altered autonomic tone, disrupted sleep, changed physical activity, and neuroendocrine remodeling that accompanies the illness.

The Bidirectional and Allostatic Load Model

The most scientifically credible current position is that the relationship is bidirectional and context-dependent, shaped by:

• Depression subtype: Melancholic/psychotic depression may genuinely involve primary HPA hyperactivation; atypical depression may involve secondary HPA hypofunction following chronic allostatic load.
• Illness stage: Acute onset episodes may involve cortisol surges; chronic, recurrent depression may show blunted HPA reactivity due to receptor downregulation.
• Genetic vulnerability: Polymorphisms in GR genes (NR3C1), FK506-binding protein (FKBP5), and CRH receptor genes modulate HPA sensitivity and interact with early adversity.
• Life stress exposure: Early-life adversity programs a sensitized HPA axis that responds to later stressors with exaggerated cortisol output, increasing depression risk.

The Glucocorticoid Cascade Hypothesis

One influential framework is the glucocorticoid cascade hypothesis (Sapolsky), which proposes that cumulative lifetime glucocorticoid exposure progressively impairs hippocampal negative feedback regulation, creating a self-amplifying cycle: stress → cortisol elevation → hippocampal damage → impaired HPA regulation → further cortisol elevation. Over decades, this cascade could convert acute stress reactivity into chronic HPA dysregulation and late-life depression.

---

Low Cortisol and Depression: The Blunted Hypothesis

The popular conception of depression as a "high cortisol" disorder is oversimplified. A significant body of evidence supports a low cortisol phenotype in specific depression presentations, and understanding this distinction is critical for cortisol MDD researchers.

Atypical Depression

Atypical depression — characterized by mood reactivity, hypersomnia, leaden paralysis, hyperphagia, and rejection sensitivity — has long been associated with blunted HPA reactivity rather than the hyperactivation seen in melancholic depression. Several studies have found:

• Lower 24-hour urinary cortisol
• Blunted CAR
• Attenuated cortisol responses to CRH stimulation
• Lower late-evening cortisol

This HPA hypofunction in atypical depression may reflect a distinct neurobiology more closely aligned with inflammatory or autonomic models than the classic glucocorticoid excess hypothesis.

Chronic and Treatment-Resistant Depression

Patients with chronic, treatment-resistant depression (TRD) frequently show blunted cortisol profiles. One interpretation is HPA exhaustion: years of stress-driven HPA activation have depleted the system's reserve capacity, leading to receptor downregulation and reduced adrenal responsiveness — sometimes described as "adrenal fatigue" in popular literature, though this specific term lacks clinical validation.

The 2025 Mendelian Randomization Finding in Context

The 2025 MR finding of lower morning cortisol associated with depression (β = −0.107) is consistent with the blunted cortisol hypothesis being a causal downstream consequence of depression — or at minimum, a heritable vulnerability factor that co-occurs with depression liability in the population.

Importantly, this does not mean elevated cortisol is irrelevant. The MR finding uses genetic instruments for population-level depression risk and may not capture the acute, severe, or melancholic presentations where cortisol elevation is most robustly documented. These findings likely reflect different populations, different time points in illness course, and different depression subtypes.

Clinical Takeaway

Clinicians and researchers should resist the assumption that "depression = high cortisol." The direction of HPA dysregulation in a given patient depends critically on:

• Depression subtype (melancholic vs. atypical)
• Illness chronicity (acute vs. chronic/recurrent)
• Comorbid trauma or PTSD history
• Time of cortisol sampling
• Current antidepressant status

---

Cortisol in Adolescents and Young Adults With Depression

The depression cortisol study findings in adolescents and young adults deserve particular attention because this population represents both a critical developmental window and a priority for early intervention.

Why the Adolescent Period Is Particularly Relevant

Adolescence involves dramatic HPA axis remodeling. Puberty alters glucocorticoid receptor expression, HPA sensitivity, and the neuroanatomical substrates of stress regulation. The developing hippocampus and prefrontal cortex are particularly vulnerable to glucocorticoid excess during this period, with implications for lifelong neurodevelopmental trajectories.

Early-onset depression during adolescence is associated with more chronic, recurrent illness and greater neurocognitive burden in adulthood — potentially mediated, in part, by HPA dysregulation during a sensitive developmental period.

What the 2024 Meta-Analysis Found

As noted above, the 2024 meta-analysis found:

• Elevated basal cortisol in depressed adolescents and young adults vs. controls (SMD = 0.87, 95% CI 0.43–1.31)
• Elevated cortisol stress reactivity (SMD = 0.68, 95% CI 0.31–1.05)
• Strong positive correlation between morning/afternoon cortisol and depression scores (r = 0.82)

The magnitude of the cortisol-depression correlation (r = 0.82) is notably larger than what is typically found in adult samples. This may reflect:

1. Reduced biological heterogeneity in younger samples (fewer comorbidities, shorter illness duration)
2. Stronger HPA reactivity in adolescents generally (the "stress sensitization" hypothesis)
3. Closer temporal proximity to precipitating stressors (less time for HPA adaptation)
4. Methodological differences across the adolescent vs. adult literature

Implications for Early Intervention

If elevated cortisol in depressed adolescents is a marker of acute HPA hyperactivation and/or a predictor of illness severity, it could theoretically inform:

• Staging of depression severity
• Selection of interventions that modulate HPA activity (exercise, CBT, mindfulness, pharmacotherapy)
• Monitoring of treatment response
• Identification of youth at highest risk for chronic/recurrent illness

However, none of these applications is currently supported by sufficient prospective evidence to warrant routine clinical implementation. Longitudinal research tracking cortisol trajectories through adolescent depression treatment is urgently needed.

---

Cortisol, Suicidal Behavior, and Severe Depression

The 2025 meta-analysis and case-control study on depression with suicidal behavior — finding elevated cortisol relative to healthy controls — adds an important clinical dimension to cortisol psychiatric research.

The Biological Logic

Suicidal behavior, particularly impulsive and acute suicidality, is associated with extreme psychological distress, which would logically activate HPA stress circuitry. Several plausible mechanisms connect cortisol elevation to suicidal risk:

• Acute cortisol surges may impair prefrontal inhibitory control, reducing the capacity to suppress suicidal impulses in vulnerable individuals.
• HPA hyperactivation is associated with CRH hypersecretion, and elevated CSF CRH has been documented in suicide decedents in postmortem studies.
• Glucocorticoid effects on serotonin signaling may reduce serotonergic tone in ways that increase impulsivity and hopelessness.
• Sleep disruption associated with nighttime cortisol elevation amplifies suicidal ideation in the context of hopelessness.

What the Research Shows

The 2025 findings align with a broader literature suggesting that suicidal depression represents a high-arousal, high-cortisol subtype of depression, in contrast to the withdrawn, low-arousal, low-cortisol phenotype of chronic atypical or burnout-related depression. This is consistent with clinical observation: many patients in acute suicidal crisis report overwhelming internal distress, agitation, and inability to sleep — a high-HPA activation phenotype.

Research and Clinical Limitations

Cortisol is not — and is nowhere near — a validated clinical biomarker for suicide risk. The research is promising but highly preliminary, and no clinical guideline recommends cortisol measurement for suicide risk assessment. The 2025 findings must be replicated in prospective cohorts with standardized cortisol protocols before any clinical translation is warranted.

---

---

Can Cortisol Tests Diagnose Depression?

This is one of the most frequently asked questions in clinical practice, and the honest answer is: not currently, and probably not in isolation, though this remains an area of active investigation.

The History: The Dexamethasone Suppression Test

In the early 1980s, there was tremendous enthusiasm for the DST as a diagnostic biomarker for endogenous depression. The test appeared to show high specificity but modest sensitivity: about 40–50% of patients with severe melancholic depression showed DST non-suppression, compared to a much smaller fraction of healthy controls.

However, subsequent meta-analyses found that:

• DST non-suppression was not specific to depression; it also occurred in dementia, anorexia, alcohol dependence, and other conditions.
• Sensitivity was too low for it to function as a diagnostic screening test.
• Its clinical utility was primarily as an index of HPA dysregulation in severe illness, not as a diagnostic criterion.

The DST was largely abandoned as a routine clinical tool by the late 1980s, though research interest in HPA biomarkers never entirely disappeared.

The Current State: Biomarker Research, Not Clinical Tool

Contemporary cortisol depression research is not attempting to develop cortisol as a standalone diagnostic — researchers are instead exploring cortisol as:

1. A stratifying biomarker: Identifying HPA-hyperactive vs. HPA-hypoactive depression subtypes that may respond differently to treatment.
2. A treatment response marker: Tracking whether cortisol normalizes with effective antidepressant treatment.
3. A relapse predictor: Using cortisol patterns to identify patients at risk for recurrence.
4. A component of multivariate biological signatures: Combining cortisol with inflammatory markers, neuroimaging, genetics, and clinical variables to improve prediction.

The 2019 meta-analysis found cortisol predicted MDD onset/relapse/recurrence with OR 1.294 (95% CI 1.035–1.616) — statistically significant but clinically modest, and confounded as the authors noted. This is not a diagnostic precision instrument; it is a population-level risk signal.

Hair Cortisol: A Promising Research Tool

One methodological advance that has revitalized cortisol biomarker research is hair cortisol analysis. Because hair grows approximately 1 cm per month, a 3-cm hair sample provides a 3-month retrospective cortisol profile — essentially an HPA "calendar" that captures cumulative stress exposure rather than a single moment in time.

Several studies have found elevated hair cortisol in depression, and this measure is far less susceptible to acute sampling confounds (time of day, current stress, food intake) than plasma or salivary cortisol. Hair cortisol may ultimately prove more useful as a research biomarker of chronic HPA dysregulation in MDD than acute cortisol measures.

---

Cortisol as a Predictor of Relapse and Treatment Response

Perhaps the most clinically actionable application of cortisol mood disorder research is its potential role in predicting treatment response and relapse in MDD.

Predicting Treatment Response

Several lines of evidence suggest that baseline HPA dysregulation may predict differential response to antidepressants:

• Patients with severe HPA hyperactivation (elevated CAR, DST non-suppression) may respond differently to SSRIs vs. tricyclics vs. psychotherapy vs. combined approaches.
• High-cortisol patients may be candidates for HPA-targeted augmentation strategies (see below).
• Failure of cortisol to normalize with antidepressant treatment may predict incomplete remission and increased relapse risk.

The 2019 meta-analysis finding that cortisol predicted MDD onset/relapse/recurrence (OR 1.294, 95% CI 1.035–1.616) is relevant here, even if effect sizes are modest. In the context of complex, multi-factorial relapse prediction models, a biological variable with even modest predictive power could add incremental clinical value.

Predicting Relapse

The cortisol literature on relapse prediction aligns with the more general hypothesis that residual HPA dysregulation after clinical remission from depression represents a vulnerability marker. Patients who achieve symptomatic remission but continue to show elevated CAR or DST non-suppression may be at higher risk for relapse than those who show full HPA normalization.

This is consistent with the broader concept of "scar" effects in depression: the biological changes induced by a depressive episode do not necessarily resolve fully with symptom resolution, and residual HPA dysregulation may represent one such scar.

Practical Research Directions

For this to move toward clinical utility, research needs to:

1. Standardize cortisol sampling protocols in treatment trials
2. Distinguish HPA normalization from symptomatic remission as distinct endpoints
3. Conduct adequately powered prospective trials with cortisol-stratified randomization
4. Develop validated composite biomarker signatures rather than relying on cortisol alone

---

Antidepressants and Cortisol: What Happens to HPA Activity?

The cortisol antidepressant interaction is a mechanistically important and clinically relevant area of HPA psychiatric research. Do antidepressants work partly by normalizing HPA axis function? The evidence is suggestive but not conclusive.

SSRIs and HPA Normalization

Selective serotonin reuptake inhibitors (SSRIs) — the most commonly prescribed antidepressants globally — have been shown in multiple studies to:

• Reduce CAR in patients with elevated pre-treatment cortisol
• Improve DST suppression in non-suppressors
• Reduce CRH gene expression in animal models
• Upregulate glucocorticoid receptor function, potentially enhancing negative feedback sensitivity

Whether these HPA changes are a cause of antidepressant response or a consequence of clinical improvement (via reduced distress, improved sleep, and restored activity) remains unclear. Probably both mechanisms operate simultaneously and are difficult to disentangle.

Tricyclic Antidepressants (TCAs)

TCAs have a longer history of HPA research behind them. Early studies consistently found that clinical response to TCAs was associated with restoration of DST suppression. Some researchers have interpreted this as evidence that TCAs work at least partly through HPA axis normalization, though this remains a hypothesis.

Mifepristone (RU-486): A Direct HPA Target

Perhaps the most direct test of the cortisol depression mechanism hypothesis is the clinical use of mifepristone — a glucocorticoid receptor antagonist — in depression treatment. Several controlled trials have examined mifepristone in psychotic depression in particular, with some positive results (rapid antidepressant effects, particularly on psychotic symptoms), though results have been inconsistent across trials.

If blocking glucocorticoid receptor signaling can rapidly reduce depression — even partially — this supports the view that GR-mediated effects of cortisol are actively maintaining depressive pathology in at least some patients.

Other HPA-Targeting Approaches Under Investigation

• CRH-1 receptor antagonists: Multiple pharmaceutical candidates targeting CRH signaling have been tested in clinical trials, with mixed results. NBI-30775 and related compounds showed initial promise but have not achieved regulatory approval.
• Ketoconazole and metyrapone: Adrenal cortisol synthesis inhibitors that reduce cortisol production; studied in small trials as augmentation strategies.
• DHEA supplementation: Counters some cortisol effects and has shown modest antidepressant effects in older adults.
• Mindfulness-based interventions and CBT: Psychotherapy, particularly mindfulness-based cognitive therapy (MBCT), has been shown to normalize CAR in some studies, providing a non-pharmacological HPA-targeting approach.

---

Cortisol and Anxiety vs. Major Depression: An Important Distinction

The 2024 pooled analysis finding — cortisol significantly associated with anxiety (OR 1.16) but not with major depression (OR 1.02) in pooled data — deserves careful consideration because it challenges assumptions embedded in much of the clinical literature.

Why Might Cortisol Be More Robustly Linked to Anxiety?

Several theoretical frameworks support a stronger cortisol-anxiety link:

1. Threat appraisal and HPA activation: The HPA stress response is specifically engaged by perceived threat and unpredictability — the core cognitive features of anxiety. Depression, by contrast, involves resignation and helplessness, which may actually dampen HPA reactivity in some models.

1. Autonomic nervous system overlap: Anxiety is associated with sympathetic nervous system activation, which parallels and drives HPA responses. The autonomic withdrawal seen in melancholic depression is a different physiological state.

1. Comorbidity confounding: Depression and anxiety co-occur in 50–60% of patients. Many studies nominally examining "depression" are actually examining a mixed anxious-depressive syndrome where the cortisol elevation may be primarily driven by the anxiety component.

1. Heterogeneity of MDD: The null association between cortisol and MDD in the pooled analysis may reflect the cancellation of a positive signal (from high-cortisol melancholic/anxious depression) by a negative signal (from low-cortisol atypical/chronic depression).

Implications for Research Design

This finding argues strongly for:

• Stratifying by anxiety comorbidity in all cortisol-MDD research
• Using dimensional rather than categorical diagnostics when studying cortisol-mood relationships
• Measuring anxiety severity as a covariate or moderator rather than treating it as noise
• Considering whether HPA hyperactivation is a transdiagnostic feature of anxious psychopathology rather than a depression-specific phenomenon

---

Clinical Implications and the Road Ahead

Where the Field Currently Stands

The HPA axis depression field is in a productive but transitional phase. The initial enthusiasm for simple hypercortisolism models has been replaced by a more nuanced appreciation of:

• Subtype-specific HPA profiles (high-cortisol melancholic vs. low-cortisol atypical phenotypes)
• Bidirectional causal relationships between depression and cortisol dysregulation
• The importance of developmental timing (adolescence vs. adulthood vs. late life)
• The potential value of hair cortisol and CAR as more robust research measures
• The strong anxiety-cortisol link that may confound many depression studies

What Is Still Unknown

Despite decades of research and a rapidly growing evidence base, several fundamental questions remain unresolved:

1. Can cortisol (or any HPA biomarker) reliably stratify MDD patients into biologically distinct subtypes with different treatment responses? This is the "precision psychiatry" goal, but the evidence base is not yet there.

1. Does normalizing cortisol through pharmacological means improve long-term depression outcomes? The mifepristone and CRH antagonist literature is suggestive but inconclusive.

1. What is the mechanistic chain linking elevated cortisol to specific neural circuit dysfunctions in MDD? Hippocampal and prefrontal cortex effects are well-documented in animal models but are harder to establish causally in humans.

1. How do inflammatory markers interact with cortisol in mediating depression? The cortisol-inflammatory interface — where glucocorticoid resistance may produce paradoxical immune activation despite high cortisol — is a rapidly developing area.

1. Can CAR or hair cortisol measurements be standardized sufficiently for multicenter clinical trial use?

The Path Forward

The most promising directions for cortisol depression research over the next decade include:

• Large, well-powered longitudinal cohort studies with prospective cortisol measurement using standardized protocols, tracking from adolescence through adulthood
• Biomarker-stratified clinical trials that randomize patients based on HPA phenotype to different treatment arms
• Multivariate biological models combining cortisol with inflammatory markers, neuroimaging, genetics, and actigraphy data
• Causal inference methods (including Mendelian randomization and natural experiments) to better characterize direction of causality
• Global collaborative consortia that pool raw data to enable adequately powered subgroup analyses by depression subtype, age, sex, and antidepressant exposure

---

Frequently Asked Questions

Does depression raise cortisol levels?

Not universally. Some forms of depression — particularly melancholic, psychotic, and severe depression — are associated with elevated cortisol and HPA hyperactivation. However, atypical depression and chronic depression are associated with blunted or low cortisol profiles. A 2025 Mendelian randomization study found that depression was actually associated with lower morning cortisol, suggesting the relationship is more complex than a simple "depression raises cortisol" model. The direction and magnitude of cortisol change depends heavily on depression subtype, illness stage, and measurement protocol.

Can cortisol tests diagnose depression?

Not currently. No cortisol-based test — including the historical Dexamethasone Suppression Test — has demonstrated sufficient sensitivity and specificity to function as a diagnostic instrument for depression. Cortisol measurements are research tools and may eventually serve as stratifying or predictive biomarkers, but they are not part of standard clinical diagnostic protocols for MDD.

Is high cortisol a cause of depression or a result of it?

The evidence points toward a bidirectional relationship. High cortisol from chronic stress can contribute to depression through mechanisms including hippocampal damage, monoamine dysregulation, and sleep disruption. But the 2025 Mendelian randomization data also support the hypothesis that depression itself causes lower morning cortisol, suggesting significant reverse causality. The relationship likely varies by depression subtype, illness stage, and individual biological vulnerability.

Are morning cortisol and nighttime cortisol different in depression?

Yes, significantly. Morning cortisol (particularly the cortisol awakening response) may be elevated, blunted, or normal depending on depression subtype. Nighttime cortisol is typically very low in healthy individuals but may be elevated in severe melancholic depression, reflecting failure of normal nocturnal HPA suppression. These two measures have different regulatory mechanisms and different clinical significance.

Do antidepressants lower cortisol?

Some evidence suggests that effective antidepressant treatment — particularly with SSRIs — normalizes HPA activity, reducing elevated cortisol in some patients. This normalization may be both a mechanism of antidepressant action (via glucocorticoid receptor upregulation) and a consequence of clinical improvement. The cortisol antidepressant interaction is an active area of research, and some pharmacological candidates specifically targeting HPA pathways (such as mifepristone) are under investigation.

Can cortisol predict treatment response or relapse?

Preliminary evidence from the 2019 meta-analysis (OR 1.294 for cortisol predicting MDD onset/relapse/recurrence) suggests cortisol has modest predictive value for illness course. Persistent HPA dysregulation after clinical remission may mark residual biological vulnerability to relapse. However, this has not yet translated into validated clinical prediction tools.

Is low cortisol ever linked to depression?

Yes. Atypical depression, burnout-related depression, depression comorbid with PTSD, and chronic/treatment-resistant depression are all associated with blunted or low cortisol profiles. The 2025 MR finding of lower morning cortisol in depression adds population-level genetic evidence supporting this association. Low cortisol depression likely reflects a distinct pathophysiology from the high-cortisol melancholic phenotype.

What is the HPA axis and how does it relate to depression?

The HPA axis depression relationship is central to biological psychiatry. The HPA axis (hypothalamic-pituitary-adrenal axis) is the brain-body circuit governing cortisol production in response to stress. In depression, this circuit is frequently dysregulated — showing either excessive activation (elevated cortisol, failed dexamethasone suppression) or insufficient activation (blunted cortisol awakening response, low diurnal cortisol). Because the HPA axis modulates brain regions critical to mood, cognition, and sleep, its dysregulation is thought to mediate some of the core features of depressive illness.

---

Summary and Key Takeaways

The field of cortisol and depression clinical research has matured substantially, but it remains rich with unresolved questions. Here are the most important conclusions from the current evidence base:

1. Cortisol dysregulation is real in MDD, but it is not unidirectional. Both hypercortisolism and hypocortisolism are documented in depression, reflecting genuine biological heterogeneity rather than inconsistent research.

1. The 2024 meta-analysis in adolescents shows some of the strongest cortisol-depression correlations in the literature (r = 0.82), suggesting the relationship is particularly tractable in younger populations.

1. The 2025 Mendelian randomization finding challenges the traditional causal model, supporting the hypothesis that depression causes blunted morning cortisol rather than vice versa — at least at a population level.

1. Cortisol is more robustly linked to anxiety than to MDD in pooled analyses, arguing for careful diagnostic stratification in research and possible reconceptualization of cortisol as a transdiagnostic internalizing disorder marker.

1. Antidepressants appear to modulate HPA activity, and HPA-targeting pharmacological strategies are under active investigation, though none has achieved regulatory approval for MDD.

1. Cortisol is not a clinical diagnostic tool, but it has potential as a research stratification and prediction biomarker in properly designed studies.

1. The most pressing research needs are longitudinal, biomarker-stratified treatment trials with standardized cortisol protocols and adequate power to detect subgroup effects.

---

This post synthesizes the published research literature and is intended for educational and professional reference purposes. It does not constitute medical advice and should not be used to guide individual clinical decisions. Clinicians and researchers should consult primary sources and current clinical guidelines.

---

---