What is the evidence supporting the use of salivary diurnal cortisol (cortisol measured in saliva throughout the day) testing?

Salivary Diurnal Cortisol Testing: Evidence-Based Assessment

Salivary diurnal cortisol testing, particularly late-night salivary cortisol (LNSC), is strongly supported by evidence as a highly sensitive and specific screening test for Cushing's syndrome, with sensitivity of 92-100% and specificity of 93-100%, though it should not replace serum cortisol testing for other conditions like critical illness-related corticosteroid insufficiency. 1, 2

Primary Clinical Application: Cushing's Syndrome Screening

Diagnostic Performance

• Late-night salivary cortisol (LNSC) demonstrates excellent diagnostic accuracy with sensitivity ranging from 92-100% and specificity from 93-100% for detecting Cushing's syndrome 1, 2
• The diagnostic utility is based on the loss of normal circadian nadir of cortisol secretion in patients with Cushing's syndrome 1
• In one study of 78 patients, LNSC identified 36 of 39 patients with proven Cushing's syndrome (92% sensitivity), and when combined with 24-hour urinary free cortisol (UFC), achieved 100% sensitivity 3
• LNSC shows higher specificity than the overnight 1-mg dexamethasone suppression test (DST) and UFC, making it particularly valuable for screening 1

Optimal Testing Protocol

• At least two or three LNSC measurements are recommended to account for variability and potential cyclic hypercortisolism 1
• Sampling should occur at the patient's usual bedtime rather than strictly at midnight, as cortisol nadir is tightly entrained to sleep onset 1
• Multiple sequential LNSC measurements are particularly useful for detecting cyclic Cushing's syndrome, where patients exhibit weeks to months of normal cortisol secretion interspersed with episodes of cortisol excess 1

Age-Specific Cut-Off Values

Research has established specific cut-off levels for excluding hypercortisolism at 10:00 PM 4:

• Ages 6-10 years: 1.0 µg/L (specificity 100%, sensitivity 87.5%)
• Ages 11-15 years: 1.7 µg/L (specificity 100%, sensitivity 100%)
• Ages 16-20 years: 1.6 µg/L (specificity 100%, sensitivity 76.2%)
• Ages 21-60 years: 1.6 µg/L (specificity 100%, sensitivity 90.9%)

Physiological Basis and Correlation with Serum

Free Cortisol Measurement

• Salivary cortisol represents unbound (free) cortisol, which is the biologically active fraction 1
• Salivary cortisol shows an exponential relationship with serum free cortisol (R²=0.903), accurately reflecting the free circulating fraction 5
• An excellent correlation exists between salivary and plasma cortisol measurements (r=0.805) 6
• Salivary cortisol may be a better index of adrenal status than total plasma cortisol, as it confirmed hypercortisolism in only half of patients with elevated total plasma levels, indicating that total plasma cortisol can be falsely elevated due to increased cortisol-binding globulin 6

Diurnal Rhythm Assessment

• Normal subjects exhibit significant diurnal variation with morning levels of 3-19 µg/L dropping to <1-2.9 µg/L at 10:00 PM 4
• In Cushing's disease patients, this circadian rhythm is absent, providing the diagnostic basis for LNSC testing 4
• Parallel diurnal rhythms are observed in both saliva and serum, with values progressively lowering from morning to evening (P<0.0001) 5

Critical Limitations and Contraindications

When NOT to Use Salivary Cortisol Testing

For critical illness-related corticosteroid insufficiency (CIRCI), salivary cortisol testing is NOT recommended over serum cortisol testing 1, 7:

• No study has demonstrated that using salivary cortisol to diagnose CIRCI leads to improved patient outcomes 1
• In one study of 57 septic shock patients, 19 were excluded due to insufficient saliva production or blood contamination, raising serious feasibility concerns 1
• The test requires enzyme immunoassay which may not be routinely available at most centers 1

Confounding Factors That Invalidate Results

Patients must avoid the following to ensure accurate results 1, 8:

• Night-shift work or disrupted sleep-wake cycles: Testing should not be performed in these patients 1
• Vigorous exercise: Must abstain for at least 24-48 hours before collection, as exercise-induced cortisol elevations can persist this long 8
• Food consumption: Patients should refrain from eating for several hours before collection, particularly avoiding sugar-rich foods which affect salivary composition 8
• Alcohol: Avoid for at least 12 hours before testing 8
• Topical hydrocortisone preparations: Can contaminate samples and cause false-positive results 1

Additional Confounding Variables

Multiple pre-analytical variables significantly impact measurements 7:

• Time of day, mental stressors, oral hygiene practices
• Collection method and materials
• Gender, age, and saliva volume 1

Analytical Methodology Considerations

Mass Spectrometry vs. Immunoassay

• Liquid chromatography-tandem mass spectrometry (LC-MS/MS) provides superior sensitivity and can detect both cortisol and cortisone, avoiding contamination from topical hydrocortisone 1, 5
• However, mass spectrometry shows better sensitivity at the expense of reduced specificity compared to immunoassay 1
• LC-MS/MS methods allow quantitation down to 39.1 pg/mL for cortisol with imprecision of 5.5-9.5% 5

Collection Device Reliability

• The Salivette® collection device provides comparable results to direct spitting with better precision (CV<1.0% vs. CV<13.0%) 5

Practical Implementation for Cushing's Syndrome Screening

Patient Preparation Instructions

Provide patients with the following specific instructions 1, 8:

1. Avoid vigorous physical activity for 48 hours before collection
2. Refrain from eating for several hours before bedtime collection
3. Avoid alcohol for 12 hours before testing
4. Do not use topical corticosteroid preparations
5. Collect sample at usual bedtime (not necessarily midnight)
6. Obtain at least 2-3 samples on different nights

When to Suspect Cushing's Syndrome

Screen high-risk populations with LNSC 3:

• Patients with poorly controlled diabetes (2-3% prevalence of unsuspected Cushing's)
• Hypertensive patients (0.5-1% prevalence)
• Patients with incidental adrenal masses (6-9% prevalence)
• Unexplained osteoporosis with vertebral fractures (11% prevalence) 2

Interpretation Algorithm

1. If LNSC is below age-specific cut-off: Cushing's syndrome is effectively excluded 4
2. If LNSC is mildly elevated (just above upper limit of normal): Consider mild Cushing's syndrome; repeat testing and consider additional confirmatory tests 1
3. If LNSC is markedly elevated: Proceed with additional diagnostic workup including UFC and DST 3
4. If initial LNSC is normal but clinical suspicion remains high: Perform multiple sequential measurements to detect cyclic hypercortisolism 1

Comparison with Other Diagnostic Tests

LNSC offers distinct advantages over other screening tests 1:

• Higher specificity than DST and UFC
• Non-invasive and stress-free collection at home
• Not affected by cortisol-binding globulin (CBG) variations that confound DST interpretation
• More practical than midnight serum cortisol which requires hospitalization or venipuncture at night

However, LNSC should be used as part of a comprehensive diagnostic approach, as combining LNSC with UFC achieves 100% sensitivity for Cushing's syndrome detection 3.