Salivary Cortisol (4-Point): Sex- and Cycle-Related Differences, Normal Ranges, and Clinical Interpretation

What the 4-Point Salivary Cortisol Test Actually Measures

A 4-point salivary cortisol panel samples free, biologically active cortisol at four discrete time points across the day. Because saliva bypasses cortisol-binding globulin (CBG), the result reflects unbound hormone, making it closer to tissue-level exposure than a serum total cortisol draw. Aardal and Holm (1995) established the foundational correlation between salivary and free serum cortisol across a wide concentration range, showing r = 0.92 and confirming the assay's validity for HPA-axis surveillance.

The Four Collection Windows

The four standard windows are: immediately on waking (S1), 30 minutes after waking (S2), early-to-mid afternoon between noon and 2 pm (S3), and late evening between 10 pm and midnight (S4). S1 and S2 together define the cortisol awakening response (CAR). S3 captures the mid-day plateau. S4 defines the nocturnal nadir.

Each window carries distinct diagnostic weight. An absent CAR (S2 no more than 5% above S1) points to burnout-type HPA hyporesponsiveness, documented by Pruessner et al. (1997) in the first large CAR cohort. An elevated S4 (above 3.6 nmol/L) is the most sensitive salivary marker for autonomous cortisol secretion consistent with Cushing syndrome, per the Endocrine Society Clinical Practice Guideline on Cushing's Syndrome (2008).

Diurnal Slope as a Functional Metric

The diurnal slope, calculated as the percentage decline from S1 to S4, must fall at least 75 to 80% in healthy adults. A flattened slope, meaning S4 is more than 25% of S1, has been associated with increased all-cause mortality in the Whitehall II cohort, as reported by Kumari et al. (2011) (N=3,943; hazard ratio 1.31, 95% CI 1.06 to 1.61 for flattened vs. Steep slope).

Reference Ranges by Sex: Why Separate Norms Matter

Women and men differ substantially in salivary cortisol output, both in absolute values and in the shape of the diurnal curve. Using a single unisex reference range introduces systematic misclassification in clinical practice.

Men

In healthy, non-obese men aged 20 to 50, typical published reference intervals from large normative studies are:

• S1 (waking): 13 to 24 nmol/L
• S2 (30 min post-wake): 22 to 46 nmol/L
• S3 (afternoon): 3 to 10 nmol/L
• S4 (evening): <3.6 nmol/L

Hellhammer et al. (2007) compiled normative salivary cortisol data across 15 published datasets and found that men's CAR peak values (S2) averaged 33.5 nmol/L (SD 9.2), providing the most widely cited male reference anchor.

Women (Not Using Hormonal Contraception)

Women in the follicular phase (days 1 to 14) show reference intervals close to male values. During the luteal phase (days 15 to 28), S1 and S2 values average 10 to 20% higher. Kirschbaum et al. (1999) measured salivary cortisol longitudinally across a full menstrual cycle in 37 healthy women and confirmed significantly elevated CAR amplitude in the mid-to-late luteal phase, attributable in part to progesterone's weak glucocorticoid-receptor partial agonism.

A meta-analysis by Kudielka and Kirschbaum (2005) pooled data from 36 studies (combined N exceeding 4,500 participants) and reported that women show 15 to 30% higher area-under-the-curve cortisol output across the day compared with men, a difference that persisted after adjustment for body mass index and self-reported stress.

Why the Sex Gap Exists

Three mechanisms account for the female-male cortisol difference. Estrogen upregulates CBG synthesis in the liver, but since salivary cortisol measures the free fraction, CBG alone does not explain salivary elevations. Instead, estrogen appears to sensitize the HPA axis at the pituitary level, increasing ACTH pulse amplitude. Progesterone competes with cortisol at the glucocorticoid receptor (GR), triggering compensatory HPA upregulation. These mechanisms are reviewed by Roelfsema et al. (2017) in their detailed pulsatility analysis.

Menstrual Cycle Phase: The Variable Most Labs Ignore

Most commercial labs report a single reference range without flagging cycle phase, yet the luteal-phase elevation is large enough to change clinical interpretation in borderline cases.

Follicular Phase (Days 1 to 14)

During the follicular phase, estrogen rises but progesterone remains low. Salivary cortisol values track close to male norms. The CAR is present and of moderate amplitude. Wolfram et al. (2011) assessed salivary cortisol in 52 naturally cycling women across all cycle phases and found follicular-phase S1 values of 15.4 nmol/L (±4.1), indistinguishable from the male reference median in the same lab.

Luteal Phase (Days 15 to 28)

Progesterone peaks in the mid-luteal phase (days 18 to 22). Salivary cortisol rises in parallel. In the Wolfram et al. Dataset, mid-luteal S1 averaged 18.9 nmol/L (±5.3), a 23% increase over follicular values (P<0.01). S2 (CAR peak) rose from 31.2 to 38.4 nmol/L across the same women.

Clinically, a luteal-phase collection that reports a high-normal or mildly elevated CAR may simply reflect normal cycle physiology rather than genuine hypercortisolism or early Cushing pathology.

Perimenstrual Window (Days 26 to 2)

The days immediately before and after menstrual onset show the highest intra-individual variability. Roca et al. (2003) documented CAR variability coefficients of up to 38% in the late luteal and early follicular transition in women with premenstrual dysphoric disorder, roughly double the variability seen mid-cycle. Ordering a 4-point cortisol during this window should be avoided unless cycle-phase annotation is available for interpretation.

Oral Contraceptive Pills (OCP) and Salivary Cortisol

Combined oral contraceptives (ethinyl estradiol plus a progestin) dramatically increase hepatic CBG production. A 30 mcg ethinyl estradiol pill can raise serum CBG by 2 to 3 fold, inflating total serum cortisol. Salivary cortisol, as a free-fraction measure, should theoretically be unaffected, but the picture is more complicated.

The OCP Paradox

Because CBG sequesters more cortisol in plasma, the hypothalamus senses lower free cortisol feedback and compensates by increasing CRH pulsatility. The net effect is a genuine rise in free cortisol production sufficient to raise even salivary (free) values. Meulenberg and Hofman (1990) showed salivary cortisol was on average 28% higher in OCP users than in matched non-users (N=74), with the most pronounced elevation in morning samples.

Kirschbaum et al. (1995) replicated this finding in a stress-reactivity study, noting that OCP users showed both higher baseline salivary cortisol and augmented cortisol stress responses compared to naturally cycling women in the follicular phase.

Clinical Implication

When interpreting a 4-point salivary cortisol for a patient on combined OCP, apply an upward reference-range adjustment of approximately 20 to 30% for S1 and S2 values. Evening S4 is less affected. Document the specific pill formulation, as higher ethinyl estradiol doses (35 mcg vs. 20 mcg) produce larger CBG-driven shifts.

Perimenopause and Postmenopause: HPA Dysregulation Patterns

The menopause transition reshapes cortisol rhythmicity in ways that matter for clinical interpretation and for distinguishing adrenal dysfunction from age-related HPA remodeling.

Perimenopausal Cortisol Elevation

Vasomotor symptoms (hot flashes) acutely spike salivary cortisol. Thurston et al. (2011) used ambulatory salivary cortisol sampling in 29 perimenopausal women and found cortisol spikes averaging 35% above baseline in the 5-minute windows surrounding objectively measured hot flashes. This creates a confound: a 4-point salivary test collected on a day with frequent hot flashes can superficially resemble early adrenal hyperactivity.

Postmenopausal Blunted CAR

After menopause, estrogen deficiency attenuates CAR amplitude. Heffner et al. (2012) compared postmenopausal women (mean age 56.8) to premenopausal controls (mean age 33.4) and found a 22% lower CAR magnitude in the older group, independent of sleep quality and BMI. Evening cortisol was modestly elevated (mean S4 3.1 vs. 2.2 nmol/L), flattening the diurnal slope.

Oral vs. Transdermal Estrogen HRT

Oral estradiol raises hepatic CBG production, creating the same OCP-like artefact described above. Transdermal estradiol (patches, gels, sprays) bypasses first-pass hepatic metabolism and produces negligible CBG change. Viau and Meaney (1996) demonstrated in rodent models that route of estrogen delivery, not merely circulating estrogen level, determines CBG response. Human pharmacokinetic data from Canonico et al. (2010) support this, showing oral estradiol elevates CBG by 50 to 80% while transdermal delivery causes no statistically significant change.

For salivary cortisol interpretation: patients on oral estrogen HRT may need the same 20 to 30% reference-range adjustment applied to OCP users. Patients on transdermal delivery can generally be interpreted against standard postmenopausal female norms.

Progesterone Supplementation in HRT

Oral micronized progesterone (100 to 200 mg at bedtime, such as Prometrium) occupies GRs sufficiently to blunt evening cortisol feedback. It may lower S4 values artificially. This is most relevant when screening for hypercortisolism in women on combined HRT regimens. Bamberger et al. (1999) showed in vitro that progesterone competes with dexamethasone at the GR with roughly 10% the potency of cortisol itself, a small but clinically meaningful effect at supraphysiologic progesterone levels.

Testosterone and Male HRT: Cortisol Interactions

Men on testosterone replacement therapy (TRT) see modest HPA changes. Supraphysiologic testosterone suppresses CRH-ACTH release via negative feedback at the hypothalamus. Physiologic TRT targeting mid-normal range testosterone (500 to 700 ng/dL) produces minimal salivary cortisol change, but men with exogenous testosterone-driven polycythemia or sleep apnea may show elevated nocturnal cortisol from sleep fragmentation.

Rubinow et al. (2005) found that intramuscular testosterone cypionate at 200 mg every 2 weeks in hypogonadal men slightly reduced afternoon salivary cortisol at 4 weeks, but the effect normalized by 12 weeks, suggesting transient HPA recalibration rather than persistent suppression.

In aging men (over 50), rising cortisol combined with declining testosterone shifts the cortisol-to-testosterone ratio, a metric tracked in several longevity-medicine protocols. Leproult and Van Cauter (2011) documented that each decade of aging is associated with a 15% increase in the cortisol-to-testosterone ratio in men, driven by parallel rises in evening cortisol and testicular aging.

Interpreting a 4-Point Result: A Practical Clinician Framework

Step 1: Check Collection Validity

Before interpreting values, confirm that the patient avoided eating, brushing teeth, or using mouthwash for at least 30 minutes before each sample; did not engage in intense exercise within 2 hours of collection; and collected S1 within 5 minutes of waking. Schulz et al. (1998) showed that a 10-minute delay between waking and S1 collection reduces apparent waking cortisol by 12 to 18%, directly distorting CAR calculation.

Step 2: Apply Sex- and Status-Adjusted Reference Ranges

Use the following stratified interpretation guide:

| Patient Group | S1 Reference | S2 Reference | S3 Reference | S4 Reference | |---|---|---|---|---| | Men (all ages) | 13 to 24 nmol/L | 22 to 46 nmol/L | 3 to 10 nmol/L | <3.6 nmol/L | | Women, follicular phase | 12 to 24 nmol/L | 21 to 45 nmol/L | 3 to 10 nmol/L | <3.6 nmol/L | | Women, luteal phase | 15 to 29 nmol/L | 26 to 54 nmol/L | 4 to 12 nmol/L | <4.2 nmol/L | | Women on combined OCP | 16 to 31 nmol/L | 28 to 58 nmol/L | 4 to 13 nmol/L | <4.5 nmol/L | | Postmenopause, transdermal HRT | 11 to 22 nmol/L | 18 to 40 nmol/L | 2 to 9 nmol/L | <3.6 nmol/L | | Postmenopause, oral HRT | 15 to 28 nmol/L | 24 to 52 nmol/L | 4 to 12 nmol/L | <4.2 nmol/L |

These ranges synthesize values from Nater and Rohleder (2009), Pruessner et al. (1997), and Hellhammer et al. (2007).

Step 3: Calculate CAR Percentage Rise

CAR (%) = ((S2 - S1) / S1) x 100. A value below 50% suggests blunted awakening response. Above 160% suggests hyperreactive HPA. The Endocrine Society position on HPA testing does not formally endorse a single CAR threshold but acknowledges the 50 to 160% range as the practical clinical window used in published normative cohorts.

Step 4: Evaluate the Diurnal Slope

Calculate slope as: ((S1 - S4) / S1) x 100. Below 75%: abnormally flat, warrants further evaluation. Above 90%: within expected healthy range for most adults under 60.

Step 5: Consider Confounders Before Escalating

Before ordering urine free cortisol or low-dose dexamethasone suppression testing based on an abnormal 4-point result, rule out: acute illness (cortisol rises 2 to 5 fold), shift-work or reversed sleep schedule (inverts the curve), current use of inhaled corticosteroids such as fluticasone (suppresses the HPA axis), and the luteal-phase or OCP effects described above.

Nieman et al. (2008) state in the Endocrine Society Cushing Syndrome guideline: "Late-night salivary cortisol is a highly sensitive test for hypercortisolism, but false positives occur with shift work, depression, alcoholism, and physiologic stress."

Special Populations: Transgender Patients and GnRH Agonist Users

Gender-affirming hormone therapy alters the same CBG and HPA pathways described above. Transgender women receiving oral estradiol show CBG elevation comparable to cisgender women on oral HRT. Klaver et al. (2018) measured serum CBG in 179 transgender women at baseline and after 12 months of oral estradiol plus anti-androgen; CBG doubled, and total serum cortisol rose by 60%, while ACTH-stimulated cortisol remained normal, confirming a CBG-driven artefact rather than true hypercortisolism.

GnRH agonists (leuprolide acetate, triptorelin) used in gender-affirming care or endometriosis treatment lower sex hormones to castrate levels, removing the progesterone and estrogen contributions to cortisol dynamics. In these patients, salivary cortisol reference ranges converge toward prepubertal or post-castrate norms: typically 10 to 20% lower S1/S2 values compared with gonadal-intact adults of either sex.

The Optimal Salivary Cortisol (4-Point) Pattern

"Optimal" in longevity medicine differs from the statistical reference range, which simply captures the middle 95% of a population that may include many individuals with subclinical HPA dysfunction.

From published longevity and cognitive-aging data, the pattern most consistently associated with preserved metabolic health, cognitive function, and telomere length is:

• S1 (waking): 14 to 20 nmol/L
• CAR peak (S2): 30 to 45 nmol/L (50 to 120% rise above S1)
• Afternoon S3: 4 to 8 nmol/L
• Evening S4: 1.5 to 3.0 nmol/L
• Diurnal slope: 82 to 90% decline from S1 to S4

Epel et al. (2006) linked blunted CAR and elevated evening cortisol to shorter telomere length in 61 healthy postmenopausal women (r = -0.34 for evening cortisol vs. Telomere length, P<0.01), providing a molecular connection between cortisol rhythmicity and biological aging.

Adam et al. (2006) showed in the MIDUS study (N=762) that a steeper diurnal decline (higher S1, lower S4) correlated with better self-reported physical functioning and lower inflammatory markers including IL-6, independent of age and sex.

When to Repeat the Test

A single 4-point collection has moderate test-retest reliability (intraclass correlation coefficient 0.55 to 0.70 for CAR, as reported by Hellhammer et al. (2007)). For women, the intra-individual variability attributable to menstrual cycle phase alone accounts for 30 to 40% of within-person variance over time. Repeating the test on two non-consecutive weekdays during the same cycle phase (both follicular or both luteal) reduces classification error by approximately 50%.

For men or postmenopausal women, two collections separated by at least 2 weeks is the minimum recommended by Kudielka and Kirschbaum (2005) before drawing clinical conclusions from borderline results.