HPA Axis Dysfunction: Causes, Symptoms, Diagnosis, and Treatment

The direct answer

HPA axis dysfunction covers two clinically opposite problems that share one anatomical circuit. Cortisol deficiency (adrenal insufficiency, including autoimmune Addison's disease and steroid-induced secondary insufficiency) causes fatigue, weight loss, and low blood pressure, and can progress to a fatal adrenal crisis if untreated. Cortisol excess (Cushing's syndrome, most often from a pituitary adenoma or from exogenous steroid use) causes central weight gain, muscle wasting, and metabolic disease. Both are diagnosed with specific hormone tests, not symptom checklists, and "adrenal fatigue" as a standalone diagnosis is not recognized in any current diagnostic classification. Vague fatigue with normal cortisol testing is a different clinical problem that should not be treated with glucocorticoids.

Naming the parts: what these terms actually mean

The hypothalamic-pituitary-adrenal (HPA) axis is a feedback loop, not a single gland. The hypothalamus releases corticotropin-releasing hormone (CRH), which signals the anterior pituitary to release adrenocorticotropic hormone (ACTH), which signals the adrenal cortex to make cortisol. Cortisol then suppresses CRH and ACTH release, closing the loop.

Three failure patterns matter clinically, and confusing them leads to the wrong test:

• Primary adrenal insufficiency (Addison's disease): the adrenal glands themselves are damaged and cannot respond to ACTH.
• Secondary (or tertiary) adrenal insufficiency: the adrenal glands are intact, but the pituitary or hypothalamus is not driving them adequately, most often because of long-term exogenous steroid use suppressing the axis.
• Cushing's syndrome / Cushing's disease: excess cortisol from any source. "Cushing's disease" specifically means a pituitary adenoma that secretes ACTH autonomously; "Cushing's syndrome" is the broader umbrella term that also includes adrenal tumors and exogenous steroid excess.

These are not points on a single severity scale. A patient can move from normal, to suppressed, to insufficient (in the low direction), and a separate patient can move from normal to subclinical excess to overt Cushing's (in the high direction). The workup and treatment plan depend on which direction is in play, which is why this page separates them rather than treating "HPA dysfunction" as one disease.

Primary adrenal insufficiency (Addison's disease)

Addison's disease occurs when the adrenal cortex cannot produce adequate cortisol, and usually aldosterone, regardless of ACTH drive. Autoimmune destruction of the adrenal cortex, targeting the enzyme 21-hydroxylase, is the dominant cause of primary adrenal insufficiency in resource-rich settings; other causes include adrenal hemorrhage, tuberculosis (still relevant in endemic regions), adrenal metastases, and congenital adrenal hyperplasia.

Symptoms include fatigue, weight loss, nausea, and low blood pressure. Skin hyperpigmentation is specific to primary disease: when cortisol is chronically low, ACTH rises, and ACTH's precursor molecule (pro-opiomelanocortin) also generates melanocyte-stimulating hormone, darkening skin creases, gums, and sun-exposed areas.

Diagnosis. A low morning serum cortisol supports the diagnosis, and a level well above the normal range makes it unlikely; values in between are indeterminate and require a dynamic test. The standard-dose ACTH stimulation test (cosyntropin) is the recommended dynamic test in current Endocrine Society guidance for both primary and secondary adrenal insufficiency; a peak cortisol that fails to rise adequately after stimulation is diagnostic. The exact numeric cutoff used to define "adequate" depends on the cortisol assay a given lab uses, which is an underappreciated source of discordant results between institutions; readers should ask their clinician which cutoff their lab applies rather than relying on a single universal number. Plasma ACTH helps distinguish primary from secondary disease: it is characteristically elevated in Addison's disease and low or inappropriately normal in secondary insufficiency. Adrenal autoantibody testing (anti-21-hydroxylase) is positive in most autoimmune cases and can reduce the need for adrenal imaging when clearly positive.

Treatment. Hydrocortisone, generally 15 to 25 mg per day split across two or three doses, with the largest dose taken on waking to approximate the normal cortisol peak, is the standard glucocorticoid replacement. Fludrocortisone replaces aldosterone in primary insufficiency. Patients should carry medical identification and an emergency injectable hydrocortisone kit, and should be taught individualized sick-day dosing rules by their endocrinologist for fever, vomiting, or surgery. Adrenal crisis, untreated cortisol deficiency causing vascular collapse, is a medical emergency; observational data suggest a meaningful case-fatality rate per episode, though the exact figure varies across cohorts and should not be quoted as a fixed universal statistic. Anyone with known adrenal insufficiency who develops persistent vomiting, severe weakness, or fainting needs emergency evaluation, not a wait-and-see approach.

Secondary adrenal insufficiency

Here the adrenal glands are structurally normal but under-stimulated because pituitary ACTH output is insufficient. Because aldosterone is regulated mainly by the renin-angiotensin system rather than ACTH, secondary insufficiency usually spares aldosterone, so hyperkalemia and hyperpigmentation are typically absent, an important bedside distinguishing feature from Addison's disease.

Causes. The most common cause worldwide is exogenous glucocorticoid use, oral, inhaled, or high-potency topical, taken for more than roughly three to four weeks. This suppresses hypothalamic CRH and pituitary ACTH output, and abrupt discontinuation can precipitate a crisis before the axis recovers. Pituitary tumors, pituitary surgery or radiation, traumatic brain injury, infiltrative disease, and postpartum pituitary infarction (Sheehan's syndrome) are other causes.

Testing and treatment. The ACTH stimulation test is still the standard first-line test, though clinicians managing patients within the first several months after pituitary surgery may prefer other dynamic tests because ACTH reserve can be transiently unpredictable during recovery. Glucocorticoid replacement mirrors Addison's disease; mineralocorticoid replacement is not needed. Axis recovery after stopping long-term steroids is variable and can take many months, and it should be tracked with serial morning cortisol or repeat stimulation testing rather than assumed on a fixed timeline.

Cushing's syndrome and Cushing's disease

Cushing's syndrome is the clinical state produced by sustained cortisol excess from any source. Cushing's disease specifically means a pituitary adenoma driving that excess, and it accounts for a majority of endogenous (non-medication) Cushing's syndrome cases, though exact proportions vary by cohort and referral pattern.

Recognizing it. The most discriminating features are centripetal weight gain, wide purple striae, proximal muscle weakness, facial plethora, easy bruising, and hypertension. Endocrine Society and European consensus guidance is consistent on one point worth stating plainly: no single symptom or single test reliably rules Cushing's syndrome in or out, which is why a structured, multi-test approach is used rather than one lab value.

Incidental adrenal masses found on imaging done for unrelated reasons are increasingly common, and a meaningful minority of these show autonomous cortisol production without overt Cushing's features, a category discussed below as subclinical autonomous cortisol secretion.

Biochemical diagnosis. Three first-line screening tests are generally used: the 1 mg overnight dexamethasone suppression test, late-night salivary cortisol (checking that cortisol fails to fall to its normal midnight nadir), and 24-hour urine free cortisol. Guideline practice generally calls for at least two abnormal results, ideally from two different test types, before pursuing source localization, because false positives occur with acute illness, depression, alcohol use, and high-dose biotin supplementation (biotin can interfere with certain cortisol immunoassays). Once cortisol excess is confirmed, plasma ACTH separates ACTH-dependent disease (pituitary adenoma or an ectopic tumor) from ACTH-independent disease (an adrenal tumor suppressing pituitary output).

Finding and treating the source. Gadolinium-enhanced pituitary MRI identifies an adenoma in a majority, but not all, confirmed Cushing's disease cases; when imaging is negative or equivocal, inferior petrosal sinus sampling can confirm a pituitary source before surgery. Transsphenoidal surgery is the recommended first-line treatment for Cushing's disease and produces remission in most microadenoma cases, though recurrence and treatment failure both occur and require follow-up options: repeat surgery, radiation, bilateral adrenalectomy, or medical therapy. Several drugs are used for cortisol excess with different regulatory status: some, like pasireotide and mifepristone, carry FDA approval for defined Cushing's indications, while others, such as ketoconazole for this use, are used off-label. Bilateral adrenalectomy controls hypercortisolism reliably in surgery-refractory disease but carries a real long-term risk of Nelson's syndrome (pituitary tumor progression after adrenalectomy removes cortisol's feedback), which is a tradeoff patients and endocrinologists weigh explicitly, not an incidental footnote.

Subclinical HPA dysregulation and the "adrenal fatigue" question

"Adrenal fatigue" is not a diagnosis recognized in ICD-11, DSM-5, or Endocrine Society guidance. The adrenal glands do not gradually wear out from ordinary life stress while showing normal biochemistry; either cortisol production is measurably abnormal, or it is not. That said, three genuinely different clinical situations get lumped under "low energy, possible adrenal issue," and they call for different next steps. The table below is a practical triage rule, not a substitute for a clinician's evaluation.

Exceptions and gray zones. A cortisol value that lands in the indeterminate range on a morning draw, or a single mildly abnormal screening test in someone acutely ill, does not sort cleanly into any row above; it means repeat testing under better conditions, not immediate labeling. Patients recovering from recent pituitary surgery or recent long-term steroid tapering also do not fit this table well because their biochemistry is in transition rather than at steady state.

There is growing interest in whether chronic low-grade inflammation contributes to symptoms that resemble subtle HPA dysregulation without meeting insufficiency or Cushing's criteria, an idea explored in reviews of inflammatory depression linking cytokine signaling to HPA axis output (PubMed, 2026 review). This is an area of active research rather than an established clinical pathway, and it should not be used to justify hormone treatment in the "normal biochemistry" row above.

Special situations

Critical illness and sepsis. Severe illness can suppress the HPA axis through inflammatory cytokine interference with CRH and ACTH signaling, and clinicians sometimes use stress-dose glucocorticoids in septic shock as part of guideline-directed critical care. A 2026 comparative study examined low-dose glucocorticoids combined with continuous blood purification in sepsis, reporting effects on inflammatory markers and T-lymphocyte subsets (PubMed, 2026); this is a single study in a specific combined-treatment context, not a general statement about glucocorticoid use in all septic patients, and the exact protocol and outcome measures should be checked before citing this to a patient or in clinical decision-making.

Post-viral fatigue. Interest in HPA recovery after infections such as COVID-19 has grown, but claims about how commonly cortisol runs low after infection, and by how much, vary across published cohorts and are not settled enough to support empirical cortisol supplementation without formal testing. Fatigue after an infection is common and usually reflects several overlapping processes, not isolated adrenal failure.

Pregnancy. Cortisol-binding globulin rises in pregnancy, raising total cortisol while free cortisol rises more modestly, so standard non-pregnant cortisol reference ranges do not apply directly. Cushing's syndrome in pregnancy is rare but carries meaningful maternal and fetal risk and needs specialist management; the exact diagnostic cutoffs used in pregnancy differ from non-pregnant ranges and should be confirmed with a lab experienced in obstetric endocrinology.

Children. Congenital adrenal hyperplasia from 21-hydroxylase deficiency is the most common cause of primary adrenal insufficiency in children and is now caught through newborn screening for 17-hydroxyprogesterone in the United States. Treatment uses weight-based hydrocortisone dosing to balance cortisol replacement against androgen suppression, and dosing should be individualized by a pediatric endocrinologist rather than estimated from adult protocols.

How clinicians typically sequence the workup

1. History first. Document every glucocorticoid exposure: oral, inhaled, topical, and injected. High-dose inhaled or topical steroids can suppress the axis measurably even without an obvious "steroid pill" history.
2. Morning serum cortisol, drawn early (commonly 7 to 9 AM). A clearly low value suggests insufficiency; a clearly high-normal or elevated value makes insufficiency unlikely; values in between require dynamic testing.
3. ACTH stimulation test when the morning cortisol is indeterminate or insufficiency is otherwise suspected.
4. Plasma ACTH, drawn with a low cortisol sample, to separate primary from secondary disease.
5. If Cushing's is suspected, two of the three first-line screening tests (dexamethasone suppression, salivary cortisol x2, urine free cortisol), avoiding interpretation during acute illness, depression, heavy alcohol use, or high-dose biotin use.
6. Source localization with pituitary MRI or adrenal imaging once biochemical hypercortisolism and its ACTH-dependence are established.

Monitoring during treatment

Patients on long-term hydrocortisone replacement are typically monitored for bone density, blood pressure, and fasting glucose over time, since even physiologic-dose glucocorticoid replacement is not risk-free. Some centers use timed cortisol measurements after the morning dose (a "cortisol day curve") to check that replacement is neither under- nor over-dosed. Patients who go into remission after Cushing's disease surgery often pass through a period of secondary adrenal insufficiency, because the previously suppressed contralateral adrenal tissue needs time to recover, and they typically need temporary hydrocortisone coverage until a repeat stimulation test confirms recovery; the recovery timeline varies by patient and should be tracked individually rather than assumed.

What is established, what is plausible, and what is not established

Established: the ACTH stimulation test and dexamethasone suppression test are the accepted first-line dynamic tests for insufficiency and excess respectively; primary insufficiency is distinguished from secondary insufficiency by ACTH level and by preserved versus impaired aldosterone/hyperpigmentation; exogenous steroid use is a common cause of secondary insufficiency; "adrenal fatigue" as an independent disease entity is not recognized by current diagnostic classifications.

Plausible but not settled: the precise epidemiologic rates of primary insufficiency and Cushing's syndrome vary meaningfully across published cohorts and should be treated as ranges, not fixed facts; the role of chronic inflammation in producing HPA-axis-adjacent symptoms without meeting insufficiency or Cushing's criteria is an active research area; post-viral cortisol changes are observed in some cohorts but their clinical significance for symptom management is unresolved.

Not established: that fatigue, salt cravings, or poor sleep with normal cortisol testing represents a form of adrenal gland exhaustion that responds to glucocorticoid supplementation. Treating this group empirically with steroids risks causing the very axis suppression the treatment claims to fix.

This paragraph is the core takeaway: HPA axis dysfunction is diagnosed by direction (too little versus too much cortisol) and confirmed with specific dynamic testing, not by symptom pattern alone; the ACTH stimulation test and the dexamethasone suppression test remain the accepted first-line dynamic tests for insufficiency and excess respectively, though exact numeric cutoffs are assay-dependent and should be confirmed against the performing lab; and no current diagnostic classification recognizes "adrenal fatigue" as a standalone disease, so normal biochemistry with real symptoms warrants investigation of other causes rather than empirical steroid treatment.