Hypothalamic-pituitary-adrenal axis
(regulation, feedback, and the circadian rhythm of cortisol)

The hypothalamic-pituitary-adrenal axis is the main neuroendocrine circuit responsible for converting internal and external stimuli into coordinated modulation of adrenocorticotropic hormone (ACTH) secretion and cortisol production by the adrenal cortex. Unlike other endocrine axes, the hypothalamic-pituitary-adrenal (HPA) axis must reconcile two apparently opposing requirements: maintaining a stable and safe glucocorticoid tone for everyday survival while simultaneously generating rapid and proportionate responses to the demands placed on the body. The result is a system that does not operate in a linear or static manner, but through multilevel dynamics in which the circadian rhythm, ultradian pulses, multistage feedback, stress, and energy status shape an adaptive output.

Understanding the HPA axis requires moving beyond the simple sequence of corticotropin-releasing hormone, ACTH, and cortisol and recognising that the glucocorticoid signal is determined not only by the amount secreted but also by its free fraction, its distribution among plasma-binding proteins, tissue prereceptor metabolism, and the sensitivity of intracellular receptors. This page examines the axis as a dynamic and temporal system, with particular emphasis on pulsatility, the circadian rhythm, the individual set point, recovery times after perturbations, and the interpretative limitations of basal tests, which often capture a single point along a trajectory rather than the true state of equilibrium.

The hypothalamic-pituitary-adrenal axis is organised as a hierarchical control circuit in which central signals converge on the hypothalamus, the pituitary gland acts as an amplifier and filter, and the adrenal gland converts a peptide input into a steroid output with broad systemic effects. The hypothalamic node is represented primarily by neurons in the paraventricular nucleus, which integrate circadian, limbic, and autonomic inputs and release corticotropin-releasing hormone (CRH) into the hypothalamic-pituitary portal system. In many conditions, CRH is accompanied by the synergistic contribution of vasopressin, which acts as a co-signal that enhances the pituitary response and reshapes the sensitivity of the system to stressors. This organisation allows the brain to determine not only how much cortisol is required, but also which temporal profile is biologically useful.

At pituitary level, ACTH is produced by corticotroph cells of the anterior pituitary through processing of the prohormone proopiomelanocortin (POMC). ACTH secretion is not a simple continuous release but is characterised by secretory bursts that are reflected in cortisol oscillations. From a control perspective, the pituitary gland responds not only to the intensity of hypothalamic drive but also to its temporal structure and to modulation by glucocorticoid feedback. This explains why the axis may appear inconsistent when assessed through isolated measurements, while displaying robust patterns when analysed longitudinally or through frequent sampling.

The adrenal gland, particularly the zona fasciculata, responds to ACTH by activating a signal-transduction cascade mediated by G protein-coupled receptors, increased cyclic adenosine monophosphate (cAMP), and activation of steroidogenic enzymes. ACTH induces both rapid effects, particularly through mitochondrial cholesterol transport regulated by the steroidogenic acute regulatory protein (StAR), and slower trophic and transcriptional effects on key enzymes. The balance between these mechanisms enables cortisol production to increase rapidly while preserving long-term secretory capacity. Adrenocortical steroidogenesis depends on the cooperation of several enzymatic steps, including those that lead to pregnenolone formation and, through specific pathways, to the final production of cortisol. This complexity requires precise regulation because persistent variations in ACTH input can remodel both the structure and responsiveness of the adrenal cortex.

The output of the axis also does not correspond exclusively to the total cortisol concentration, but to the fraction that is effectively available to tissues. A substantial proportion of circulating cortisol is bound to corticosteroid-binding globulin (CBG) and albumin, whereas the free fraction, which crosses cell membranes and activates intracellular receptors, can vary despite unchanged total cortisol concentrations when transport proteins or their binding affinity change. At tissue level, the signal is further modulated by prereceptor conversion between cortisol and cortisone through the isoforms of 11β-hydroxysteroid dehydrogenase, which can increase or reduce local glucocorticoid exposure without necessarily altering plasma values. In terms of control logic, the HPA axis is therefore not a single tap, but a system of multiple valves distributed across the brain, circulation, and peripheral tissues.

Overall, the architecture of the axis enables robust stabilisation of homeostasis and a flexible response to changing demands. Robustness derives from negative feedback and the hierarchical nature of the circuit, while flexibility derives from temporal modulation of the signal and peripheral regulators of bioavailability. This explains how the HPA axis can maintain a coherent configuration over long periods while simultaneously changing its operating mode when the biological context shifts towards acute stress, metabolic adaptation, or systemic illness.

The secretion of ACTH and cortisol is intrinsically temporal. Under physiological conditions, the axis exhibits a circadian rhythm with a morning peak and a nighttime nadir, but this profile is not a smooth wave. It is constructed from a succession of ultradian pulses that combine to generate the overall pattern. Pulsatility is not a marginal detail because many cellular responses to glucocorticoids depend on the dynamics of exposure and on changes in receptor activation over time. In other words, the physiological glucocorticoid signal is not simply the presence of cortisol, but the alternation between phases of stimulation and partial recovery, which allows tissues to interpret the message efficiently.

Ultradian pulses can emerge from mechanisms intrinsic to the pituitary-adrenal interaction, in which ACTH feed-forward stimulation of steroidogenesis and rapid cortisol feedback on the pituitary gland cooperate with the physiological delays inherent in the system. These dynamics can generate oscillations without requiring a dedicated central ultradian oscillator, demonstrating that pulsatility is an emergent property of the circuit. Within this context, the brain can modulate the amplitude and frequency of secretory bursts by acting on hypothalamic drive and co-modulatory signals, thereby achieving precise control of the output without maintaining continuously elevated activation.

The circadian rhythm of cortisol reflects integration between the central clock and the HPA axis. The suprachiasmatic nucleus and peripheral circadian systems modulate the excitability of circuits controlling the paraventricular nucleus and, consequently, the probability of CRH signal activation. The result is a profile that prepares the body for daytime activity by promoting energy availability, vascular tone, and immunometabolic readiness, while reducing glucocorticoid tone during the night, with implications for sleep, synaptic plasticity, and immune regulation.

Sleep is not merely a period of endocrine inactivity but an active modulator of the system. Sleep quality, duration, and alignment of the sleep-wake cycle influence the dynamics of the axis, with potential changes in the evening and nighttime profile and in the awakening response. Shift work, circadian misalignment, and exposure to light at inappropriate times during the night can reshape the timing of the axis and alter the relationship among perceived stress, ACTH, and cortisol. Physiologically, this means that two individuals with identical total cortisol production may display different profiles and, consequently, different biological effects, because tissues are sensitive not only to how much cortisol is present but also to when it is present.

A clinically relevant component is the cortisol awakening response and the variability of the morning peak, which can be altered by chronotype, sleep restriction, and stress. Since endocrinological testing is frequently based on morning measurements, understanding temporal dependence is essential to avoid inappropriate conclusions. The HPA axis operates within a context of complex rhythmicity. Ignoring this dimension can transform temporal physiology into an apparent disorder, particularly when modest variations observed at different times or under non-comparable sleep conditions are interpreted without adequate context.

Negative feedback from cortisol is central to the stability of the axis. The glucocorticoid reduces ACTH secretion and modulates CRH production, preventing progression towards hypercortisolism or chronic insufficiency of the signal. This control is not a single switch. It operates at hypothalamic and pituitary levels and within limbic circuits that influence drive to the paraventricular nucleus. Feedback is therefore not simply an interaction between a hormone and a gland, but a distributed control system that integrates endocrine regulation with the neural regulation of stress, emotion, and behaviour.

The molecular basis of feedback is mediated by activation of the glucocorticoid receptor and, in certain tissues, the mineralocorticoid receptor. Once activated, these receptors modulate gene transcription and the availability of mediators that regulate neuronal excitability, corticotroph sensitivity, and neuropeptide production. Differences in receptor affinity and distribution allow the system to respond to both physiological oscillations and larger variations. In particular, the presence of prereceptor systems such as type 2 11β-hydroxysteroid dehydrogenase in selected tissues protects certain mineralocorticoid receptors from activation by cortisol, enabling functional separation between glucocorticoid and mineralocorticoid signalling in specific anatomical sites.

Feedback does not operate identically at each node. The pituitary gland acts as a rapid filter of the output, modulating ACTH in relation to cortisol exposure and the availability of the hypothalamic signal. The hypothalamus integrates the glucocorticoid signal with information concerning energy status, inflammation, and the external environment. This difference explains why the axis can, under certain conditions, display apparently discordant combinations of ACTH and cortisol or alterations in the circadian profile without necessarily indicating a primary adrenal disorder. The HPA axis is a multilevel control system, and when one level changes its priorities, the global output may be remodelled while remaining consistent with an adaptive physiological logic.

Cortisol feedback is also responsible for suppression of the axis in the presence of exogenous glucocorticoids. Prolonged exposure to glucocorticoid therapy can reduce hypothalamic drive and corticotroph activity, leading to functional atrophy of the adrenal cortex and a risk of adrenal insufficiency after withdrawal or under stressful conditions. Recovery dynamics vary among individuals and depend on treatment duration, dose, the specific glucocorticoid used, and individual factors. This is a clinical example of the power of feedback. The system is designed to reduce output when the glucocorticoid signal is abundant, but this efficiency is accompanied by recovery times that may be prolonged.

Overall, negative feedback establishes the stability of the glucocorticoid set point and enables control that does not oscillate chaotically. However, precisely because feedback acts at multiple nodes and can be remodelled, the interpretation of isolated cortisol or ACTH measurements requires consideration of the recent history of the system, the presence of glucocorticoid therapy, and the circadian context in which the axis is operating.

The concept of the set point of the hypothalamic-pituitary-adrenal axis describes the tendency of each individual to maintain a relatively stable personal configuration of ACTH and cortisol, with a circadian profile and ultradian dynamics that are often more consistent within the individual than across the population. Interindividual variability is substantial because it depends on receptor sensitivity, differences in central stress circuits, sleep quality, body composition, baseline inflammatory status, and genetic factors that modulate cortisol transport and metabolism. Consequently, normality does not correspond to a single number but to an overall configuration that tends to recur over time under comparable conditions.

The biological variability of cortisol is more complex than that of many other hormones because the signal is strongly time-dependent. A substantial proportion of its fluctuations is physiological and reflects ultradian pulses and the circadian rhythm. Another component reflects modulation by stress, physical activity, sleep, and energy status. In practice, a single measurement may overestimate or underestimate the true position of the axis and, most importantly, may not be representative if obtained during a secretory pulse or a phase of circadian transition. This leads to an important interpretative principle: the most reliable information on axis function derives from reproducibility under comparable conditions or from integrated measurements, rather than from an isolated result.

The set point of the axis is also influenced by the proportion of cortisol that is biologically available. CBG is the principal plasma carrier, and during inflammation or hormonal changes its concentration and affinity may change, altering the relationship between total cortisol and the free fraction. In addition, tissue prereceptor metabolism, particularly through 11β-hydroxysteroid dehydrogenase, can increase or reduce local exposure to cortisol independently of total circulating concentrations. For this reason, two individuals with similar values may experience different biological effects, and the same individual may display changes in glucocorticoid effect without substantial variations in total cortisol when protein binding and tissue conversion change.

In clinical and physiological terms, the set point means that population reference ranges are useful but inevitably approximate tools for describing a temporal and individualised system. Correct interpretation requires integration of sampling time, sleep and stress conditions, medication therapy, and, when necessary, dynamic tests that assess the ability of the axis to respond or suppress. This leads directly to the concepts of readjustment time and dynamic memory, which are particularly evident during recovery after suppression or severe illness.

Hysteresis of the hypothalamic-pituitary-adrenal axis describes the phenomenon through which the relationship between ACTH and cortisol, or between cortisol and clinical status, may depend on the temporal trajectory that brought the system to a particular point. At the same cortisol concentration, an individual may be in a different physiological condition depending on whether they are recovering from hypercortisolism, undergoing glucocorticoid-induced suppression, or experiencing a period of persistent stress. This property is consistent with a system in which regulation is distributed and includes transcriptional adaptations, receptor remodelling, and trophic changes in the peripheral gland.

Recovery times after a substantial perturbation are determined by several components. At central level, hypothalamic drive and corticotroph sensitivity may require time to return to a basal operating mode after prolonged periods of increased feedback. At adrenal level, steroidogenic capacity and responsiveness to ACTH also depend on trophic integrity and the expression of components of the steroidogenic cascade. Chronic exposure to exogenous glucocorticoids provides the clearest example. Suppression of the circuit may be followed by slow and variable recovery, with a risk of clinically significant insufficiency, particularly during stress, surgical procedures, or infections.

The dynamic memory of the axis also becomes apparent after critical illness or severe stress. In some cases, the profile may display phases characterised by apparent dissociation between ACTH and cortisol or responses that are not fully congruent, reflecting temporary reconfiguration of the system. The physiology of the HPA axis is designed to protect the body during unstable periods. This may generate transitional states in which equilibrium is not immediately restored and basal tests are difficult to interpret without considering the chronology of events.

The endocrinological significance of hysteresis is practical because it calls for caution when relying on single measurements while the axis may be in transition. In suspected secondary or glucocorticoid-induced adrenal insufficiency, interpretation must consider the treatment history and biological recovery times. During stress or recovery from illness, it is necessary to distinguish primary dysfunction from temporary adaptation. The HPA axis is not an instantaneous sensor but a circuit with its own time constants and biological memory, and these characteristics form an integral part of its normal physiology.

The HPA axis is the principal endocrine translator of stress, but stress is not a single signal. Physical, emotional, inflammatory, and metabolic stressors activate partially overlapping circuits that converge on the paraventricular nucleus and on centres that modulate CRH and vasopressin. This architecture allows a proportionate response. The system can increase glucocorticoid output when necessary to support blood pressure, energy substrate availability, and control of inflammation, but it can also reshape the response when the primary biological requirement is energy conservation or tissue repair.

Energy status modulates the axis bidirectionally. On the one hand, cortisol increases the availability of glucose and other energy substrates and promotes metabolic adaptations that support the stress response. On the other hand, signals that inform the brain about energy balance and body composition influence the likelihood of axis activation and its temporal profile. During caloric restriction, rapid weight loss, or sleep disruption, the system may change its tone and dynamics, not necessarily in a single predictable direction, but through remodelling that reflects integration between survival requirements and the metabolic costs of glucocorticoid signalling.

Inflammation acts as both a central and peripheral modulator. Cytokines and inflammatory mediators can activate the axis but can also alter cortisol transport and metabolism, changing the relationship between production and biological availability. During severe inflammation, CBG may decrease or change its properties, favouring a larger free fraction or altered tissue distribution. Prereceptor metabolism may also reshape local exposure according to specific immune requirements. In this context, total cortisol measurements may not accurately represent the glucocorticoid effect and require a more physiological interpretation.

A crucial aspect is integration with the circadian system. The central clock and peripheral clocks coordinate the timing of the axis and tissue sensitivity to cortisol. The biological effect of the glucocorticoid therefore also depends on the time of day at which the signal occurs. Circadian misalignment, as seen with shift work and nighttime light exposure, can alter the profile and contribute to metabolic and cardiovascular consequences, not only because it changes how much cortisol is produced, but because it disrupts the coherence between the endocrine signal and the time window of greatest tissue sensitivity.

These central modulations clarify that the HPA axis is an integrative system. Its normal physiology includes the ability to change its operating mode according to context, without every variation necessarily representing primary disease. Correct endocrinological interpretation must therefore consider stress, sleep, the circadian rhythm, energy status, and inflammation as determinants of the observed profile.

During systemic illness, particularly when severe, the hypothalamic-pituitary-adrenal axis may undergo profound remodelling. Under stable conditions, the axis produces a coherent temporal profile and effective feedback. During haemodynamic instability, intense inflammation, medication exposure, and impaired perfusion, the relationship among ACTH, total cortisol, and biological effect may change. The body may increase cortisol production as a survival response, but bioavailability and receptor sensitivity may be altered, making it difficult to infer glucocorticoid activity from a single plasma parameter.

In critical care, clinical and physiological discussion includes the concept of critical illness-related corticosteroid insufficiency, in which the glucocorticoid effect may be inadequate in relation to the severity of stress. This condition does not necessarily correspond to an absolutely low cortisol concentration because its central feature is a disproportion between biological demand and effective response. In these patients, changes in CBG, altered cortisol clearance, variations in the free fraction, and tissue glucocorticoid resistance may contribute to functionally inadequate activity even when total cortisol does not appear reduced. The physiology of the system therefore becomes inseparable from clinical interpretation.

Systemic inflammation may also remodel feedback dynamics. The brain and pituitary gland operate within an environment of mediators that can alter sensitivity to cortisol and modify drive to the paraventricular nucleus. In addition, medications commonly used in critical care settings may interfere with steroid synthesis or metabolism. In this context, an apparently active axis may not ensure an adequate response in terms of blood pressure and inflammatory modulation, while an apparently normal axis may actually be undergoing a transitional phase or temporary uncoupling.

A physiological perspective is useful for understanding these phenomena. The HPA axis is designed to adapt, but adaptation may become inefficient when peripheral determinants of the signal change dramatically. Systemic illness reveals the limitations of models based on a single value and requires an approach that integrates temporal dynamics, clinical context, and knowledge of transport, metabolism, and receptor sensitivity.

Interpretation of hypothalamic-pituitary-adrenal axis testing must begin with the recognition that cortisol displays marked temporal variability. Morning cortisol is frequently used as a reference point, but its usefulness depends on the circadian context, awakening time, sleep quality, and the presence of acute stress. A single value can be informative only when obtained under standardised conditions and interpreted cautiously, because an ultradian pulse or circadian transition may substantially shift the result.

A second consideration is the distinction between total cortisol and the biologically available signal. Since CBG is the principal plasma-binding protein, conditions that reduce its concentration or alter its affinity may make total cortisol measurements misleading. In these settings, assessment of the free fraction or use of measurements that more directly reflect free cortisol, such as salivary cortisol at specific time points or integrated urinary assessment, may more closely correspond to physiology. Cortisol clearance and tissue prereceptor metabolism also contribute to the separation between the concentration measured in blood and the biological effect, particularly during critical illness or marked inflammation.

Measurement of ACTH is useful when assessing the direction of central control and distinguishing primary from central forms of adrenal insufficiency. However, ACTH is also pulsatile and sensitive to stress and procedural factors, and its measurement requires careful pre-analytical handling and contextual interpretation. In many situations, physiology requires dynamic testing to assess the functional capacity of the axis rather than an instantaneous value. The cosyntropin test evaluates adrenal reserve and the ability to respond to ACTH stimulation. Other tests may assess central drive, but they require specialist settings and awareness of their associated risks.

In suspected central adrenal insufficiency, the clinical and treatment history is decisive. Suppression caused by exogenous glucocorticoids can produce results that depend on recovery time and the temporal trajectory of the axis, and a test performed too early or during a transitional phase may not represent the stable state. Physiological assessment therefore integrates clinical information, timing, dynamic behaviour, and selection of the test best suited to the clinical question, avoiding the misclassification of normal variability or adaptive transition as disease.

When interpreting the physiology of the axis, three concepts reduce interpretative errors and clarify the significance of test results:

• time, because ACTH and cortisol follow ultradian and circadian patterns and an isolated measurement may not describe the steady state;
• biologically active fraction, because CBG and prereceptor metabolism can dissociate total cortisol from its tissue effect;
• context, because sleep, stress, medications, and systemic illness remodel the axis and its feedback mechanisms without necessarily indicating primary adrenal disease.

In summary, interpretation of hypothalamic-pituitary-adrenal axis testing requires respect for the nature of the system: a temporal and adaptive control circuit, with signals that change throughout the day, plasma binding that modifies bioavailability, and feedback that introduces memory and recovery times. This perspective does not diminish the value of testing, but improves its accuracy by placing each result within the real physiology of the HPA axis.