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

Why the Diurnal Curve Matters More Than a Single Cortisol Value

A single morning serum cortisol can miss both hypercortisolism and hypocortisolism if the sample timing is off. The 4-point salivary test measures free (unbound) cortisol at four standardized time-points, giving a curve rather than a snapshot.

Salivary cortisol tracks free serum cortisol with a 2 to 3-minute lag and correlates with free serum cortisol at r = 0.92 in healthy adults across multiple validation studies [1]. Because saliva bypasses cortisol-binding globulin (CBG), the test is not distorted by CBG changes from oral estrogens, pregnancy, or liver disease in the same way a total serum cortisol would be.

The Four Time-Points and What Each Measures

Waking sample (T1). Collected within 30 minutes of rising, before eating, drinking, or brushing teeth. This value represents baseline adrenal output after the overnight nadir.

Cortisol awakening response (CAR). Some protocols add a second waking sample 30 to 45 minutes after T1 to quantify the CAR, a rapid ACTH-driven surge that can rise 50 to 160% above T1. Blunted CAR appears in burnout, PTSD, and adrenal insufficiency [2].

Noon sample (T2). Reflects mid-morning decline. Values should be roughly 50 to 60% of the waking sample in healthy adults.

Late-afternoon sample (T3). Collected around 4 to 5 PM. Continued decline confirms an intact diurnal slope.

Bedtime sample (T4). The most diagnostically sensitive point for hypercortisolism. The Endocrine Society's 2016 Clinical Practice Guideline on Cushing's syndrome states: "We suggest the use of late-night salivary cortisol (LNSC) as an initial screening test for Cushing's syndrome, with two measurements recommended to improve diagnostic accuracy" [3].

Reference Ranges Across Major Laboratories

Reference intervals vary by assay platform. The Endocrine Society guideline [3] and published validation data [4] converge on these approximate values:

| Time-point | Typical Range (nmol/L) | Typical Range (ng/dL) | |---|---|---| | Waking (T1) | 13 to 24 nmol/L | 471 to 870 ng/dL | | Noon (T2) | 5 to 10 nmol/L | 181 to 362 ng/dL | | 4 to 5 PM (T3) | 3 to 8 nmol/L | 109 to 290 ng/dL | | Bedtime (T4) | <3.6 nmol/L | <130 ng/dL |

Always compare results against the specific lab's own reference interval. The table above reflects commonly cited ranges, not universal cut-offs.

Normal and Optimal Ranges: What "Good" Looks Like

"Normal" and "optimal" are not the same target. A normal range encompasses the middle 95% of a reference population; an optimal range reflects values associated with the best health outcomes in prospective cohort data.

The Diurnal Slope as a Health Marker

A steep morning-to-bedtime slope, often quantified as the cortisol slope index, predicts lower inflammatory burden and lower all-cause mortality. A flattened slope (small difference between waking and bedtime values) was associated with a hazard ratio of 1.26 for all-cause mortality over 12 years in the English Longitudinal Study of Ageing (ELSA, N = 3,695) [5]. That finding persists after adjusting for smoking, BMI, and cardiovascular disease.

Optimal Waking Values

Most longevity-medicine practitioners use a waking T1 target of 15 to 22 nmol/L (roughly 540 to 800 ng/dL). Values below 13 nmol/L in a symptomatic patient prompt evaluation for adrenal insufficiency. Values above 24 nmol/L at waking, combined with an elevated bedtime T4, raise concern for autonomous cortisol secretion and warrant follow-up with a 1 mg overnight dexamethasone suppression test per Endocrine Society guidance [3].

The Bedtime Cut-Off

A bedtime salivary cortisol above 7.5 nmol/L (the commonly cited 2 standard deviation upper limit used in several validation studies) carries a sensitivity of approximately 92 to 100% and specificity of 93 to 100% for Cushing's syndrome in most published series [4]. For subclinical HPA overdrive, even values of 4 to 7 nmol/L at bedtime, combined with a blunted diurnal slope, may warrant clinical attention.

How Medications Shift the 4-Point Curve

This is the core clinical problem: a medication can produce a curve pattern that mimics disease, masks disease, or genuinely reflects HPA suppression or activation. Systematic documentation of all drugs, including topical and inhaled agents, before interpreting any cortisol panel is mandatory.

Glucocorticoids: The Major Suppressors

Oral prednisone and prednisolone. Any dose of systemic glucocorticoid suppresses ACTH via negative feedback, flattening or obliterating the diurnal curve. Prednisone 5 mg/day for as few as 7 days suppresses the morning cortisol awakening response measurably [6]. Patients on chronic low-dose prednisone (even 2.5 mg/day) may show waking salivary cortisol values below 5 nmol/L, mimicking primary adrenal insufficiency. The distinction matters: in exogenous suppression, DHEA-S is also suppressed, and an ACTH stimulation test will show blunted response proportional to duration of suppression.

Hydrocortisone replacement therapy. Patients prescribed hydrocortisone for Addison's disease or secondary adrenal insufficiency complicate salivary testing significantly. The timing and dose of their last hydrocortisone tablet directly predicts each time-point value. A patient taking hydrocortisone 10 mg at 6 AM will show an artificially elevated T1 and T2, then a physiologic-looking decline into T3 and T4. Clinicians at HealthRX routinely ask patients to withhold the morning hydrocortisone dose until after the T1 sample to obtain a true baseline, though this must be done carefully in patients with fragile adrenal status.

Inhaled corticosteroids (ICS). High-dose ICS (fluticasone propionate 1,000 mcg/day or more, or budesonide 1,600 mcg/day or more) produce measurable HPA suppression in a significant minority of patients. A systematic review and meta-analysis (N = 3,728 patients across 37 trials) found that high-dose ICS suppressed morning cortisol by a mean of 1.74 nmol/L (95% CI: 0.55 to 2.93) compared to placebo [7]. Salivary testing in patients on high-dose ICS should document ICS dose, delivery device, and whether a spacer is used.

Megestrol acetate. This progestational agent has intrinsic glucocorticoid receptor activity. Patients taking megestrol acetate for appetite stimulation (commonly 400 to 800 mg/day in oncology settings) can develop clinical adrenal insufficiency and a fully suppressed diurnal curve [8].

Topical and intranasal corticosteroids. Often overlooked. Mometasone nasal spray at standard doses (200 mcg/day) has minimal systemic absorption; fluticasone nasal spray at 200 mcg/day was associated with measurable urinary cortisol reduction in one pediatric study but negligible suppression in most adult data [9]. Topical dermal corticosteroids applied over large body surface areas can produce suppression proportional to potency and coverage area.

Oral Contraceptives and Exogenous Estrogens

Oral contraceptives (OCs) containing ethinyl estradiol raise CBG by 2 to 3 fold, increasing total serum cortisol substantially while leaving free salivary cortisol largely unchanged [10]. This is why salivary testing is preferred over serum total cortisol in OC users: a serum total cortisol of 28 mcg/dL in a woman on ethinyl estradiol may represent a normal free fraction, whereas salivary cortisol will reflect the biologically active portion accurately.

Transdermal or vaginal estradiol (common in HRT protocols) raises CBG less dramatically than oral ethinyl estradiol, producing smaller serum cortisol elevations with minimal effect on salivary free cortisol [11].

SSRIs and SNRIs

The acute effect of SSRIs on the HPA axis is stimulatory: single-dose paroxetine or escitalopram increases cortisol in challenge paradigms. With chronic dosing (8+ weeks), the net effect is typically normalizing or suppressing, particularly for the bedtime value. A randomized controlled trial by Harmer et al. Found that 7-day citalopram pretreatment blunted cortisol responses to social stress tasks, but diurnal basal cortisol was not significantly altered in the non-depressed subjects [12]. In patients with MDD being treated with SSRIs, the expected finding is a normalization of the blunted CAR seen in untreated depression, rather than a new pathological shift.

Stimulants: Methylphenidate and Amphetamines

Stimulant medications increase sympathetic tone and can mildly raise cortisol at T1 and T2 time-points on days of use. A crossover study in adults with ADHD found that methylphenidate 30 mg increased salivary cortisol by approximately 20% in the first 2 hours post-dose [13]. Patients should note stimulant timing on their collection log; samples taken 1 to 3 hours after stimulant ingestion may over-represent the waking value.

GLP-1 Receptor Agonists

Semaglutide and liraglutide do not directly bind glucocorticoid or mineralocorticoid receptors. However, the substantial weight loss produced by these agents can secondarily normalize cortisol metabolism. Adipose tissue expresses 11-beta-hydroxysteroid dehydrogenase type 1 (11beta-HSD1), which regenerates active cortisol from cortisone. As fat mass decreases, 11beta-HSD1 activity decreases, and urinary free cortisol often falls [14]. Salivary cortisol reflects this indirectly: patients losing 10 to 15% body weight on semaglutide 2.4 mg (as seen in STEP-1, N = 1,961, mean 14.9% weight loss at 68 weeks [15]) may show modestly lower T1 values after significant weight loss, without any HPA pathology.

Opioids and Chronic Opioid Therapy

Chronic opioid use suppresses the hypothalamic-pituitary-adrenal axis through mu-receptor-mediated inhibition of CRH release. Opioid-induced adrenal insufficiency (OIAI) is underdiagnosed. Published prevalence estimates range from 8% to 29% of patients on long-term opioids [16]. The 4-point salivary cortisol in these patients typically shows a low flat curve with waking values below 10 nmol/L across all four time-points. An ACTH stimulation test distinguishes secondary HPA suppression (cortisol response preserved) from primary adrenal failure (absent response).

Anticonvulsants: Carbamazepine and Phenytoin

Both drugs are potent CYP3A4 inducers, accelerating cortisol metabolism to inactive metabolites. The net effect on salivary free cortisol is a mild reduction in measured values, despite normal or even elevated cortisol production rates [17]. This can cause falsely reassuring salivary results in patients on these anticonvulsants; a 24-hour urinary free cortisol with concurrent serum cortisol production rate assessment may be needed for complete picture.

Ketoconazole and Adrenal Enzyme Inhibitors

Ketoconazole at doses of 400 to 1,200 mg/day inhibits CYP11A1 and CYP11B1, reducing cortisol synthesis. This is used intentionally in Cushing's syndrome management. Patients on ketoconazole for any indication (including antifungal therapy at lower doses, 200 to 400 mg/day) may show dose-dependent reductions across all four time-points [18]. The Endocrine Society notes that even antifungal doses of ketoconazole can reduce cortisol output by 20 to 30%.

Collection Protocol and Pre-Analytic Confounders

Getting the biology right matters only if the sample is collected correctly. Pre-analytic errors are the single largest source of incorrect 4-point cortisol results.

Standard Collection Instructions

Patients should collect samples using the salivette or passive drool method per their laboratory's kit. The 15-minute rule applies: no food, drink (including water), tooth brushing, or vigorous exercise within 15 minutes of collection. Blood in the saliva (from gum disease or dental procedures) elevates salivary cortisol artificially, so patients with active periodontal disease should notify their clinician [19].

Timing Documentation

Each sample tube must be labeled with the exact collection time, not just "morning" or "bedtime." Cortisol falls approximately 0.5 to 1.0 nmol/L per 30-minute delay in the waking sample due to the steep early-morning decline. A T1 collected at 7:00 AM versus one collected at 8:30 AM in the same patient can differ by 3 to 5 nmol/L through physiology alone, not pathology.

Stress and Illness

Acute psychological stress and intercurrent illness both spike cortisol via hypothalamic CRH release. The T1 sample should be rescheduled if the patient has had fever, a significant emotional event, or intense physical exertion the prior evening. A minimum of two separate collection days is recommended before clinical decisions are made, consistent with the Endocrine Society's recommendation for duplicate late-night salivary cortisol measurements in Cushing's screening [3].

Interpreting the Curve Pattern: A Decision Framework

Recognizing the pattern before acting on it prevents unnecessary workup and missed diagnoses.

Pattern 1: Flat Low Curve (All Four Points Below 8 nmol/L)

This pattern suggests adrenal insufficiency (primary or secondary), exogenous glucocorticoid suppression, or chronic opioid use. Check ACTH, DHEA-S, and a full medication list before ordering stimulation testing. If the patient is on any systemic glucocorticoid or opioid, the drug is the most likely explanation.

Pattern 2: Elevated Bedtime Only (T4 Above 7.5 nmol/L, T1 Normal)

This is the classic early Cushing's pattern. Two separate elevated late-night salivary cortisol values have sensitivity exceeding 90% for Cushing's syndrome [3]. Order a 24-hour urinary free cortisol and 1 mg overnight dexamethasone suppression test next. Rule out pseudo-Cushing states: major depression, alcoholism, and severe obesity can all produce mildly elevated late-night values.

Pattern 3: High T1, Normal Decline to T4

Usually benign, particularly in acute stress or in patients taking stimulants or short-acting SSRIs in the morning. If the patient is asymptomatic and on no suppressive or stimulating medications, repeat the panel in 4 to 6 weeks.

Pattern 4: Blunted Slope (High T1, High T4, Small Difference)

Seen in chronic psychological stress, night-shift workers, patients with poorly controlled sleep apnea, and chronic opioid users. Associated with the increased all-cause mortality risk reported in ELSA [5]. Address the underlying cause before attributing to medication.

Pattern 5: Normal T1, Blunted or Absent CAR

If the protocol includes a 30-minute post-waking sample, a CAR rise of less than 50% suggests HPA hypo-reactivity. This pattern is reported in burnout syndrome, PTSD, and early adrenal insufficiency [2]. Stimulants taken immediately after waking can blunt the CAR by attenuating the ACTH pulse.

Medication Management: When to Hold, Adjust, or Accept

Not every medication-driven cortisol shift requires action. The clinical question is: does the medication-induced pattern represent a genuine risk to the patient, or is it an expected pharmacologic effect?

When to Hold Medications Before Testing

Hydrocortisone: hold the morning dose until after T1, as described above. Do this only with physician supervision and only in patients not at risk for adrenal crisis.

Stimulants (methylphenidate, mixed amphetamine salts): consider collecting T1 before the morning dose on collection days, then resuming the normal schedule. This removes the stimulant artifact from the waking value.

Topical corticosteroids over large areas: consider a 48-hour drug-free window if feasible, documenting this on the requisition.

When to Accept the Drug Effect and Test Anyway

Patients on antiepileptic CYP inducers: accept that salivary values may be mildly low, and consider adding a 24-hour urinary free cortisol for cross-reference.

Patients on oral contraceptives: salivary cortisol is specifically preferred over serum in this group; no medication modification is needed.

Patients on chronic opioids: do not abruptly discontinue before testing. The flat-low curve is expected and clinically informative. The result guides whether an ACTH stimulation test is warranted.

Dose Adjustments Driven by the Curve

In patients receiving hydrocortisone replacement therapy, the 4-point salivary panel guides dose timing rather than total daily dose alone. A patient with an elevated T3 but low T4 may benefit from shifting the afternoon hydrocortisone dose 30 to 60 minutes later. A patient with an elevated T1 and T2 combined with a low T3 and T4 may be over-replacing in the morning and under-replacing in the afternoon. Published hydrocortisone pharmacokinetic modeling data support dose intervals of 6 to 8 hours for conventional immediate-release tablets to approximate physiologic rhythm [20].