The HPA axis is the hormonal stress pipeline: hypothalamus releases CRH, pituitary releases ACTH, adrenals release cortisol, cortisol suppresses the upstream via negative feedback. The rhythm of that cascade is as informative as the absolute level.
What cortisol does
Cortisol mobilizes energy (glucose release from liver, free fatty acids from adipose), suppresses inflammation acutely, enables the fight-flight response, and contributes to circadian anchoring (peak shortly after waking). Chronic elevation produces central adiposity, insulin resistance, muscle breakdown, hippocampal atrophy over years, and impaired immune response.
The shape of the curve matters:
- Cortisol Awakening Response (CAR): the 30-45 min post-wake rise, usually 50-100% above waking level. A blunted CAR associates with chronic stress, burnout, and cardiometabolic disease.
- Diurnal slope: the fall from morning peak to evening nadir. A flat slope (small drop) associates with cancer progression, depression, and metabolic dysfunction across multiple cohorts ( Adam et al. 2017 ).
- Nighttime nadir: should be near-zero by 11pm-2am. Elevated nighttime cortisol disrupts sleep.
What to test
Morning serum cortisol alone is usually insufficient. Misses pattern. A single 8am number can look "normal" in someone with a flat diurnal curve.
4-point salivary cortisol (waking, 30 min post-waking, early afternoon, bedtime) gives the shape. Commercially available via ZRT, Dutch Test, and similar labs for ~$150-250.
Dutch Test (dried urine total + metabolites) is the more comprehensive test for adrenal + sex hormones + melatonin. ~$350. Often over-ordered by functional medicine practitioners who then recommend supplement stacks of questionable evidence.
When to escalate to clinical testing:
- Symptoms suggestive of Cushing's (central obesity, striae, moon face, muscle wasting, severe fatigue): 24-hour urinary free cortisol or low-dose dexamethasone suppression test. Endocrinologist.
- Symptoms suggestive of Addison's (postural hypotension, hyperpigmentation, salt craving, fatigue): morning cortisol + ACTH, cosyntropin stimulation test. Urgent referral.
Cortisol normal range: reference values by test type
Cortisol reference ranges depend on the test (serum vs saliva vs urine), the time of day, and the lab. The numbers below are the most common adult ranges; check your own lab's reference column on the actual report.
| Phase | Dose | Notes |
|---|---|---|
| Serum AM (6-8am) | ~6-23 mcg/dL (165-635 nmol/L) | Peak of the diurnal curve. Single measurement; misses rhythm. |
| Serum PM (4-6pm) | ~2-12 mcg/dL (55-330 nmol/L) | Should be roughly 50% of AM. AM/PM ratio > 2 is the healthy pattern. |
| Salivary waking (within 30 min of waking) | 0.27-1.5 ng/mL | First post-waking sample; baseline of the cortisol awakening response (CAR). |
| Salivary 30 min post-waking (CAR peak) | Should rise 50-100% above waking sample | Healthy CAR is 50-100% rise. Flat CAR is the early HPA dysregulation signal. |
| Salivary afternoon (4-5pm) | 0.094-0.359 ng/mL | Mid-curve sample; should be 30-50% of waking value. |
| Salivary bedtime (10-11pm) | < 0.359 ng/mL (typically < 0.1) | Trough. Elevated bedtime cortisol is the signal of evening sympathetic drive. |
| 24-hour urinary free cortisol (UFC) | < 60 mcg/24h | Total daily output. Used to screen for Cushing's, not for fine HPA tuning. |
| Late-night salivary cortisol | < 0.13 mcg/dL (< 3.6 nmol/L) | Cushing's screen; > 0.15 mcg/dL warrants endocrinology workup. |
A single in-range AM cortisol does not mean the HPA axis is healthy. Two patterns to watch for that single tests miss:
- Flat diurnal slope: AM cortisol normal, but PM cortisol almost as high (AM/PM ratio < 1.5). Associated with chronic stress, depression, sleep debt, and metabolic syndrome. Flat slope predicts mortality independent of mean cortisol.
- Blunted CAR: 30-min post-waking sample fails to rise 50% above waking. Early signal of HPA dysregulation in shift workers, chronic burnout, post-viral fatigue, and PTSD.
If you have symptoms (fatigue, low mood, poor stress tolerance, exercise intolerance, sleep fragmentation) and a single AM serum cortisol in range, the next test is the 4-point salivary panel, not another serum draw.
"Adrenal fatigue" and what's actually going on
"Adrenal fatigue" as a medical diagnosis does not exist. The Endocrine Society has issued explicit position statements against it. What's usually going on in people told they have adrenal fatigue:
- Chronic stress + poor sleep + under-recovery from exercise: this is common and real. The label wraps a behavioral cluster in pseudo-medical framing.
- Hypothyroidism: overlap in fatigue symptoms; a full thyroid panel catches this.
- Iron deficiency anemia or B12 deficiency: catches another slice.
- Depression or anxiety disorder: large overlap in presenting symptoms.
- Low-grade infection (Epstein-Barr reactivation, Lyme subsets): sometimes real.
HPA dysregulation is real; it looks like cortisol pattern disruption, not adrenal output failure. The fix is upstream: stress reduction, sleep repair, exercise modulation, not "adrenal support" supplements.
Chronic cortisol dysregulation also shows up in aging markers: Epel 2004 (n=58) found chronically stressed caregivers had telomere lengths equivalent to 9-17 years older than matched controls ( Epel et al. 2004, n=58 ). Cortisol is not the only mediator, but it is a central one.
Correcting mild HPA patterns
| Phase | Dose | Notes |
|---|---|---|
| Sleep repair | 7-9h consistent, 16-19°C bedroom | Highest-ROI. Poor sleep flattens diurnal slope and raises nocturnal cortisol. |
| Morning bright light | 10+ min outdoor, within 60 min of waking | Anchors CAR timing; restores circadian cortisol-melatonin inverse relationship |
| Aerobic exercise, morning-midday | Zone-2 3-4x/week | Lowers chronic cortisol over 8-12 weeks; late-night cardio opposite effect |
| Caffeine timing | No caffeine within 90 min of waking | Acute cortisol boost from caffeine blunts endogenous CAR over time |
| Alcohol reduction | Zero within 4h of bed | Alcohol raises nocturnal cortisol and fragments sleep |
| Stress-reduction practice | Slow breathing or meditation 10 min/day | Vagal tone up, cortisol AUC down over 4-6 weeks |
Supplements with cortisol claims
- Ashwagandha (300-600 mg/day): 3-4 small RCTs show ~25% reduction in morning cortisol in stressed adults. Directional, not definitive. See ashwagandha entry.
- Phosphatidylserine: older data on blunting cortisol response to exercise; effect sizes small.
- Rhodiola, holy basil, adrenal glandular, licorice root: regularly marketed, minimal RCT support for specific cortisol outcomes.
Supplements layer weakly on top of behavioral changes that are 10x more impactful.
Counter-view
Functional medicine practitioners argue 4-point salivary testing + adaptogen protocols produce symptomatic improvement their patients report; some of this is probably real (stress + behavior changes during a testing episode), some is regression to the mean, some is the attention effect. The conventional endocrinology position: if you don't have Cushing's or Addison's, "cortisol" is not the lever you should be pulling; sleep, exercise, and stress behaviors are.
