Cortisol (S4) - Dinner

Your body runs on a cortisol clock. Levels should peak shortly after you wake and fall steadily through the day, reaching their lowest point near bedtime. The dinner-time saliva sample, often called S4, captures the evening leg of that decline and tells you whether your stress system is settling down on schedule.

An evening reading that stays high, or a flat slope from morning to night, has been linked in large population studies to higher cardiovascular mortality, worse mental health, and metabolic disturbance. This single sample is most useful as one point on a daily curve rather than a stand-alone diagnosis, but it is one of the easier ways to see whether your circadian rhythm is intact.

What This Sample Actually Captures

Cortisol is the main glucocorticoid hormone made by the adrenal cortex under control of the brain-pituitary-adrenal feedback loop (the hypothalamic-pituitary-adrenal axis, or HPA axis). In blood, most cortisol is bound to carrier proteins, but saliva captures only the free, biologically active fraction. That makes salivary cortisol useful for tracking the part of cortisol that actually reaches your tissues.

The S4 sample is collected around dinner time, when a healthy rhythm should be well into its evening descent. By itself, the number means less than it does when paired with the morning samples (S1 through S3) and bedtime (S5) to draw a slope. A high evening reading on a normal background suggests your cortisol is failing to wind down. A high reading with a flat day-long curve suggests the underlying rhythm itself is disturbed.

Why a High Evening Reading Matters

Heart Disease and Mortality

In a study of 2,305 adults with high blood pressure, every standard-deviation rise in midnight cortisol was linked to about 24% higher risk of new cardiovascular events (hazard ratio 1.24), while a steeper morning-to-evening drop in cortisol was protective, with about 14% lower risk per standard-deviation increase in slope steepness (hazard ratio 0.86). In the Whitehall II cohort of 4,047 civil servants, people whose cortisol stayed flat across the day had higher all-cause and cardiovascular mortality. The KORA-F3 cohort of 1,090 adults reached a similar conclusion. The pattern across these studies is consistent: a higher evening floor, or a flatter slope, tracks with worse heart outcomes.

What this means for you: the dinner-time number is most informative when read alongside your morning value. A high S4 next to a normal morning peak hints that your evening shutdown is impaired, which is the pattern these long-term studies flagged.

Mental and Physical Health Across the Board

A meta-analysis of 80 studies found that flatter daily cortisol slopes consistently tracked with worse mental and physical health, with the strongest signal for immune and inflammatory outcomes. Higher serum cortisol in midlife has also been associated with lower brain volume, more amyloid buildup, and worse delayed memory, particularly in men, in a study of 277 adults at risk for Alzheimer's disease. Most of the brain evidence comes from blood-based cortisol, not salivary dinner cortisol specifically, but the rhythm signal it reflects is the same biology.

Metabolic Disruption and Late Eating

A randomized crossover trial in 20 healthy volunteers compared eating dinner at 18:00 versus 22:00. The 22:00 dinner produced higher nocturnal glucose, delayed fat clearance, reduced fat burning, and a higher cortisol level through the night. The cortisol in that trial was measured in blood rather than saliva, so it is supportive rather than direct evidence for the dinner saliva sample, but the direction is clear. Repeated late dinners over time push your evening cortisol the wrong way.

Adrenal Disease at the Extremes

Persistent elevation of late-evening salivary cortisol is one of the standard screening signals for Cushing syndrome, a condition of cortisol excess. Late-night salivary cortisol carries sensitivity and specificity above 90% in that setting. At the other extreme, deficient cortisol points toward adrenal insufficiency, which usually shows up as a low morning value rather than an abnormal dinner reading. Most people testing themselves outside a hospital setting are looking for circadian disruption, not these uncommon adrenal diseases, but the same sample can help flag them.

What Healthy Looks Like at Dinner

There is no consensus reference range published specifically for the dinner-time S4 sample. The closest anchors come from late-night salivary cortisol studies and large population datasets, which is why the values below should be read as orientation rather than as a definitive cutoff.

These ranges come from healthy adult cohorts measured by liquid chromatography mass spectrometry (a precise lab method called LC-MS/MS) and from population-based salivary cortisol curves. They are illustrative, not universal targets. Your lab may report different numbers and units.

Sources: Larsson et al. 2009; Bäcklund et al. 2020; Ponzetto et al. 2020. Compare your results within the same lab over time for the most meaningful trend. Different methods can produce different numbers for the same saliva sample.

Why One Reading Is Not Enough

Salivary cortisol moves around a lot from day to day. In a 1,345-person multilevel analysis, single readings were substantially affected by within-person variation, and repeated sampling was needed to characterize a true diurnal pattern. Long-term stability of daily cortisol parameters across eight months sits at intraclass correlations of about 0.2 to 0.55, where 1.0 would mean perfect repeatability. Translation: a single high or low S4 sample tells you very little. The trend tells you a lot.

Plan to collect samples on at least two separate days, ideally at the same clock time and in similar circumstances. If you are making a deliberate change such as moving dinner earlier, improving sleep, or starting stress management, retest in 8 to 12 weeks to see whether your evening number is heading down. Annual rechecks are reasonable for general tracking. The most useful comparison is your slope from morning to evening on each test day, not any one number in isolation.

What to Do With an Abnormal Result

If your dinner cortisol is high once, repeat the full daily curve before drawing any conclusion. If it stays high across two or three days alongside a flat morning-to-evening slope, treat that as a circadian signal and pair it with related companion tests: a fasting glucose and HbA1c (your three-month average blood sugar), a high-sensitivity C-reactive protein (a marker of low-grade inflammation), and a lipid panel. If the elevation is striking and persistent, especially with weight gain in the trunk, easy bruising, or rising blood pressure, an endocrinologist can run a low-dose dexamethasone suppression test or 24-hour urinary free cortisol to rule out true cortisol excess. If your morning value is low and the evening value is relatively preserved, that flatter pattern is more often about lifestyle and circadian disruption than about adrenal disease.

When Results Can Be Misleading

A single reading can be thrown off by ordinary, temporary factors that shift the number without reflecting a real change in your underlying biology.

• Recent food, drink, or brushing: eating, drinking coffee, or brushing teeth within 30 to 60 minutes of collection can contaminate the saliva sample and produce a misleading reading. Wait at least 30 minutes after any of these before spitting into the tube.
• Acute stress and meal anticipation: an argument, a workout, or even smelling food before the sample can briefly raise cortisol. Caffeine in particular has been shown in randomized work to raise cortisol responses across multiple stressors.
• Combined oral contraceptives: these raise total cortisol in blood by increasing the carrier proteins cortisol binds to, but the salivary free cortisol fraction is generally less affected. Total serum cortisol on a contraceptive may look high without your true free cortisol being elevated.
• Kidney disease: chronic kidney disease blunts the daily decline of cortisol and raises evening levels. The high reading reflects altered clearance, not necessarily adrenal disease.

Drugs are also a common cause of confusing results. Antidepressants and antipsychotics tend to lower basal cortisol. Stimulant medications used for attention-deficit conditions tend to raise it. None of these mean your adrenal gland is sick, but they will shift the number you see on the report. Note any of these medications when comparing results over time.