Salivary Cortisol (4-Point) Medication-Driven Changes: What Your Diurnal Curve Reveals

The direct answer

Salivary cortisol testing measures the biologically active, unbound fraction of cortisol at four fixed times across the day, and it is preferred over a single serum draw because a flattened or reshaped curve carries diagnostic information that one number cannot. A meaningful share of curve abnormalities referred for adrenal workup are not adrenal disease at all; they are the expected pharmacologic footprint of a glucocorticoid, an opioid, a stimulant, or a hormonal medication the patient is already taking. Before ordering confirmatory testing such as an ACTH stimulation test or a dexamethasone suppression test, the medication list should be reconciled against the pattern seen, because several of the most common causes of an abnormal-looking curve are iatrogenic and reversible rather than pathologic.

What the test is, and what it is not

This is a salivary test, distinct from a single serum cortisol, a 24-hour urinary free cortisol collection, and a dexamethasone suppression test, although these are often used together. Salivary cortisol reflects free hormone and is not affected by changes in cortisol-binding globulin (CBG) the way total serum cortisol is. That distinction matters most in patients on oral estrogens, during pregnancy, or with liver disease, where CBG shifts can make total serum cortisol look abnormal when the biologically active fraction is not.

The four standard collection points are:

• Waking (T1): collected within 30 minutes of rising, before food, drink, or tooth brushing. Reflects adrenal output after the overnight nadir.
• Midday (T2): typically collected around noon; expected to be substantially lower than T1 in a healthy diurnal pattern.
• Late afternoon (T3): collected in the 4 to 5 PM window; continued decline supports an intact diurnal slope.
• Bedtime (T4): the time-point most useful for detecting hypercortisolism, since cortisol should be at its lowest near sleep onset.

Some protocols add a fifth sample 30 to 45 minutes after waking to capture the cortisol awakening response (CAR), the short ACTH-driven surge that follows sleep offset. A blunted CAR has been described in burnout, PTSD, and adrenal insufficiency in the research literature, though the CAR is not part of every commercial 4-point panel and its clinical cut-offs are less standardized than the basic diurnal curve.

Reference ranges are assay- and lab-specific. Different salivary cortisol platforms report different units and cut-offs, and a value flagged as abnormal on one lab's report may fall inside another lab's reference interval. Any interpretation should use the ordering lab's own reference range rather than a number quoted from a different assay.

Evidence boundary: what is established, what is plausible, what is not

Established: Salivary cortisol reflects free cortisol and correlates with free serum cortisol; the Endocrine Society's clinical practice guideline for Cushing's syndrome supports late-night salivary cortisol as a first-line screening tool, generally requiring two separate abnormal measurements before proceeding to confirmatory testing. Exogenous glucocorticoids (oral, and high-dose inhaled) suppress the HPA axis through ACTH negative feedback, and this is a well-described pharmacologic mechanism, not a controversial one. Chronic opioid therapy is a recognized cause of opioid-induced adrenal insufficiency through suppression of hypothalamic CRH release.

Plausible but not settled at the level of precise numbers: The exact magnitude of ICS-induced cortisol suppression, the exact percentage rise in cortisol after a stimulant dose, and the exact prevalence of opioid-induced adrenal insufficiency all vary substantially across the published studies that report them, and the specific figures attached to the source material behind this article could not be independently verified for this draft. Where a number could not be confirmed against the primary paper, it has been described in general terms below rather than stated as a precise, load-bearing figure.

Not established: That a single abnormal salivary cortisol value, in the presence of a plausible medication explanation, indicates adrenal disease. A drug effect should be considered and, where feasible, accounted for or re-tested around, before a patient is sent for stimulation or suppression testing.

How specific medication classes shift the curve

Glucocorticoids: the main suppressors

Any systemic glucocorticoid, including prednisone and prednisolone, suppresses pituitary ACTH release through negative feedback, which flattens or eliminates the normal diurnal slope. This effect is dose- and duration-dependent, and low chronic doses can still produce measurable suppression of the morning cortisol awakening response within about a week of use. In a patient with a flat, low curve, exogenous glucocorticoid suppression should always be considered before assuming primary adrenal failure; DHEA-S is typically also low in exogenous suppression, and the pattern usually resolves as the drug is tapered under medical supervision.

High-dose inhaled corticosteroids (used for asthma or COPD) can produce measurable HPA-axis suppression in a minority of patients on the highest approved doses. Standard clinical doses have a much smaller, and often negligible, effect on adult HPA function. Documenting the specific inhaled corticosteroid, its dose, delivery device, and spacer use is more informative than assuming any inhaled steroid is equivalent to any other.

Megestrol acetate, a progestin used for appetite stimulation in oncology and palliative settings, has intrinsic glucocorticoid receptor activity at typical doses and has been reported to cause clinically significant adrenal suppression, including a flat diurnal curve, in some patients.

Topical and intranasal corticosteroids at standard doses generally have minimal systemic effect on salivary cortisol, but topical steroids applied over large body surface areas, at higher potency, or under occlusion can produce suppression proportional to dose and coverage.

Hydrocortisone replacement therapy

Patients already on hydrocortisone for primary or secondary adrenal insufficiency complicate interpretation of their own salivary panel, because the timing of the last dose directly shapes the curve. A morning dose can produce an artificially high T1 and T2 followed by a decline that looks physiologic but is really just the drug wearing off. Some clinicians ask patients to hold the morning hydrocortisone dose until after the T1 sample is collected to get a truer baseline; this should only be done under direct physician supervision, because withholding replacement steroid carries a real risk of adrenal crisis in a patient who depends on it.

Oral contraceptives and exogenous estrogen

Ethinyl estradiol-containing oral contraceptives raise cortisol-binding globulin, which raises total serum cortisol substantially while leaving free salivary cortisol comparatively unaffected. This is the main reason salivary testing, rather than total serum cortisol, is generally preferred in patients on oral contraceptives, during pregnancy, or with conditions affecting CBG. Transdermal or vaginal estradiol used in hormone therapy protocols has a smaller effect on CBG than oral ethinyl estradiol and correspondingly less effect on serum cortisol interpretation.

SSRIs and SNRIs

The acute pharmacologic effect of SSRIs tends to be stimulatory to the HPA axis, while chronic dosing (roughly eight weeks or more) is generally reported as neutral to mildly normalizing for basal diurnal cortisol, particularly in patients whose baseline cortisol awakening response was blunted by untreated depression. In a patient with major depressive disorder who starts an SSRI, a normalizing CAR is the expected finding rather than a new abnormal pattern that needs separate workup.

Stimulants

Methylphenidate and amphetamine-based stimulants raise sympathetic tone and can raise cortisol at the waking and midday time-points on days the medication is taken, particularly in the first couple of hours after dosing. Collection logs should record stimulant timing precisely; a T1 sample drawn shortly after a stimulant dose may not represent a true medication-free baseline.

GLP-1 receptor agonists

Semaglutide and liraglutide do not bind glucocorticoid or mineralocorticoid receptors and have no direct pharmacologic effect on cortisol synthesis or clearance. Substantial weight loss on these agents, however, is associated with reduced adipose expression of 11-beta-hydroxysteroid dehydrogenase type 1, the enzyme that regenerates active cortisol from inactive cortisone in fat tissue. Patients who lose a large share of body weight on GLP-1 therapy may show modestly lower cortisol values over time as a secondary metabolic effect of weight loss itself, not as evidence of HPA-axis pathology, though this pathway is better documented for urinary and serum cortisol than for the salivary panel specifically.

Chronic opioid therapy

Chronic opioid use suppresses the HPA axis through mu-receptor-mediated inhibition of hypothalamic CRH release. Opioid-induced adrenal insufficiency (OIAI) is under-recognized, and published estimates of how common it is vary widely across study populations and opioid doses; a single precise prevalence figure should not be treated as settled. The typical 4-point pattern in OIAI is a low, flat curve across all four time-points. An ACTH stimulation test can help distinguish this secondary (opioid-driven) suppression, where some cortisol response is often preserved, from primary adrenal failure, where the response is largely absent. Opioids should not be abruptly discontinued purely to "clean up" a cortisol test; the flat-low curve in a chronic opioid patient is itself clinically informative and should prompt a conversation about whether adrenal insufficiency needs formal evaluation, not medication discontinuation on its own.

Anticonvulsant enzyme inducers

Carbamazepine and phenytoin are CYP3A4 inducers that accelerate cortisol metabolism to inactive metabolites. This can produce a mildly low salivary cortisol reading despite a normal, or even compensated-higher, underlying cortisol production rate, which means a "low" salivary result in a patient on one of these drugs can be falsely reassuring or falsely concerning depending on what it is compared against. A 24-hour urinary free cortisol, interpreted alongside the salivary panel, may be needed for a complete picture in these patients.

Ketoconazole and other adrenal enzyme inhibitors

Ketoconazole inhibits steroidogenic enzymes and reduces cortisol synthesis in a dose-dependent way; this effect is used intentionally as a treatment strategy in some cases of Cushing's syndrome, but it also means that ketoconazole prescribed for an unrelated antifungal indication can lower a patient's salivary cortisol values across all four time-points.

Collection errors that masquerade as biology

Pre-analytic error is a common source of an unexpected result, and it should be ruled out before a curve is treated as either drug-driven or disease-driven.

• No food, drink (including water), tooth brushing, or vigorous exercise within 15 minutes of collection.
• Blood in saliva, from gum disease or a recent dental procedure, can artificially raise measured cortisol.
• Exact collection clock time should be recorded, not a general label like "morning." Cortisol falls quickly through the early morning hours, so a T1 collected 90 minutes later than intended in the same person can produce a materially different number without any change in health status.
• A T1 sample should be rescheduled after a feverish illness, a major emotional event, or unusually intense exertion the prior evening, since acute stress and illness both raise cortisol through the same CRH-driven pathway that the test is trying to characterize.
• Current guideline practice favors two separate collection days over a single day's result when the question is Cushing's screening, precisely because day-to-day variation and collection errors are common.

Drug-or-disease decision framework

This framework is meant to be worked through before ordering confirmatory adrenal testing, not as a substitute for clinical judgment or for direct physician evaluation of a patient with concerning symptoms.

Step 1: List every corticosteroid-family and HPA-active drug the patient takes, including inhaled, topical, intranasal, and injected forms, and opioids, stimulants, hormonal contraceptives, anticonvulsant enzyme inducers, and ketoconazole. Many of these are missed because patients do not think of an inhaler or a nasal spray as a "medication."

Step 2: Match the curve shape to the most likely explanation.

Step 3: If a drug fully explains the pattern and the patient is asymptomatic, document the explanation and avoid further adrenal workup unless new symptoms emerge.

Step 4: If a drug only partially explains the pattern, or the patient has symptoms suggestive of true adrenal disease (unexplained fatigue, weight change, hypotension, hyperpigmentation, or signs of hypercortisolism), proceed to confirmatory testing rather than attributing everything to medication.

Step 5: Never stop hydrocortisone replacement, or abruptly discontinue opioids, glucocorticoids, or other HPA-active medications, solely to "clean up" a cortisol test. Any hold or adjustment of these medications for testing purposes should be directed by the prescribing clinician.

When to seek urgent care rather than wait for lab results

Symptoms suggesting an adrenal crisis, severe hypotension, confusion, persistent vomiting, or profound weakness in a patient on chronic steroids or with known adrenal insufficiency, are a medical emergency and should prompt urgent evaluation rather than waiting for a scheduled salivary cortisol result.

What this article does not establish

This article describes general pharmacologic mechanisms and reported patterns from the endocrinology literature; it does not replace an individualized evaluation, and it does not provide dosing instructions for any of the medications discussed. Several precise figures that commonly circulate in secondary summaries of this topic, exact percentage suppression from inhaled corticosteroids, exact prevalence of opioid-induced adrenal insufficiency, exact mortality hazard ratios tied to a flattened diurnal slope, could not be independently confirmed against a primary source for this draft and should be verified against the original trial or guideline before being published as a precise number.

A qualified clinician who reviews the patient's full medication list alongside the specific curve shape is best positioned to decide whether a pattern warrants confirmatory testing, a medication adjustment, or reassurance. This article is educational and does not substitute for that individualized review.