Cortisol, DHEA-S, and the adrenal panel — HPA axis interpretation

If you have ever pulled an "AM cortisol" off a routine panel, seen 16 mcg/dL, and assumed the HPA axis was fine, this article is for you. Cortisol is one of the most context-dependent values in clinical chemistry. The same number on the same person can mean wildly different things depending on what time the blood was drawn, how the previous night went, whether stimulants are on board, what training load looks like, and — most importantly — what DHEA-S is doing alongside it. A panel that does not include DHEA-S, the cortisol:DHEA-S ratio, and ideally a 4-point salivary curve simply cannot characterise HPA axis function.

This piece is aimed at biohackers running daily caffeine (often 300+ mg), modafinil cycles, racetams, or other agents that push the HPA axis, alongside high-frequency training schedules where chronic load is the norm. The framing assumes you already understand cortisol exists; the goal is to give you a working interpretation framework so a future panel actually tells you something.

The HPA axis in brief

The hypothalamic-pituitary-adrenal axis is a tiered feedback loop. The hypothalamus releases corticotropin-releasing hormone (CRH), which prompts the anterior pituitary to release adrenocorticotropic hormone (ACTH), which signals the adrenal cortex to synthesise cortisol from cholesterol via pregnenolone. Cortisol then exerts negative feedback on both the hypothalamus and pituitary, dampening further CRH and ACTH output. Under normal conditions this loop produces a strongly diurnal pattern with a sharp morning peak, a steep daytime decline, and a low overnight nadir.

Calling cortisol the "stress hormone" undersells its role. Cortisol is permissive for nearly every major physiological process: it sustains gluconeogenesis and glycogenolysis, modulates vascular tone and catecholamine sensitivity, regulates immune cell trafficking, supports memory consolidation, and gates the transition between sleep and wake states via the cortisol awakening response (CAR). Chronically low cortisol is just as dangerous as chronically high. The clinical question is rarely "is cortisol high or low?" It is "what pattern is cortisol displaying, and what does that pattern reveal about HPA regulation?"

AM serum cortisol: what it captures and what it misses

Serum cortisol drawn between 8 and 9 am captures the post-CAR plateau — the period immediately after the morning peak has subsided into the early daytime baseline. The reference window most labs use is roughly 6 – 22 mcg/dL.

There are three serious limitations to this measurement. First, timing is not negotiable. Cortisol drops fast through the morning; a 7:30 am draw and a 10:30 am draw on the same person can read 22 and 9 mcg/dL respectively, and both can be physiologically normal. If you are tracking cortisol over time, you must standardise draw time to within a 30-minute window or the trend is meaningless.

Second, serum cortisol is acutely reactive. A stressful commute, a missed meal, an argument, an early-morning fasted training session, or 400 mg of caffeine taken at 6 am can all spike the 9 am draw by 30 to 60 percent. This is not assay noise — it is real physiology that is not representative of baseline HPA function.

Third, serum cortisol measures total cortisol, the vast majority of which is bound to corticosteroid-binding globulin (CBG) and albumin. Only roughly 5 percent is biologically active free cortisol. Anything that shifts CBG — oestrogen exposure, severe illness, hepatic dysfunction — distorts total cortisol independently of HPA output.

DHEA-S: the adrenal reserve marker

Dehydroepiandrosterone sulfate (DHEA-S) is produced by the zona reticularis of the adrenal cortex and serves as a precursor to downstream androgens and oestrogens. Unlike cortisol, DHEA-S has a long half-life and minimal diurnal variation, which makes a single morning draw genuinely representative.

DHEA-S is best understood as a marker of adrenal reserve. It declines steadily with age — peak values occur in the mid-20s and drop roughly 2 percent per year thereafter — but it also drops sharply under sustained HPA stress. The mechanism is partly substrate competition: when CRH and ACTH drive sustained cortisol synthesis, the shared upstream precursor pregnenolone gets preferentially shunted toward the cortisol pathway at the expense of DHEA. This is the so-called "pregnenolone steal," and although the term is debated, the downstream pattern — high-normal cortisol with depleted DHEA-S — is robustly observed in chronically stressed subjects.

A 32-year-old man with cortisol of 17 mcg/dL and DHEA-S of 480 mcg/dL is in a different physiological state than a 32-year-old with cortisol of 17 mcg/dL and DHEA-S of 110 mcg/dL. The first has a working HPA axis. The second is showing early HPA dysregulation with depleted reserve, even though every individual marker sits within "normal" lab ranges.

The cortisol:DHEA-S ratio as an HPA stress index

Once you have both values, the ratio becomes the most useful single number on the adrenal panel. Divide AM cortisol (mcg/dL) by DHEA-S (mcg/dL). The optimal ratio sits in the 0.02 – 0.05 range. Above 0.1, the axis is biased toward catabolism. Above 0.2, you are looking at a meaningful HPA stress signal.

A high cortisol:DHEA-S ratio predicts catabolic dominance — net protein breakdown exceeding synthesis, suppressed lean mass accretion despite training, slower wound healing, and impaired immune surveillance. It is also tightly associated with altered T4 to T3 conversion; chronic cortisol elevation inhibits the type 1 and type 2 deiodinase enzymes that produce active T3, which is one of the reasons elevated cortisol shows up clinically as low-T3 fatigue. If your ratio is high, your thyroid panel deserves the same attention.

The ratio also has implications for sex hormones. Cortisol and testosterone draw from a shared steroidogenic pathway upstream of pregnenolone, and sustained cortisol demand can functionally compete with testosterone production. A man with a chronically elevated cortisol:DHEA-S ratio whose testosterone panel is trending downward despite no other obvious cause is showing exactly this pattern.

The 4-point saliva cortisol curve

Salivary cortisol measures unbound, biologically active free cortisol. It is non-invasive and can be sampled at multiple time points across a single day — typically waking, +30 minutes (the CAR peak), midday, late afternoon, and bedtime. A clean 4-point curve reveals the diurnal pattern that serum cannot capture.

The healthy pattern is sharp: a 50 to 100 percent rise from waking to +30 min (the cortisol awakening response), a steep decline through midday, a low afternoon plateau, and a near-floor reading at bedtime. Disruption of this shape is diagnostically meaningful in ways serum cannot match. A flat morning curve with no CAR rise is the signature of late-stage HPA suppression and is heavily associated with burnout, post-viral fatigue, and chronic overtraining. A normal morning peak with elevated evening cortisol points instead to acute stress, sleep debt, or late-day stimulant use. A reversed curve — low morning, high evening — is rarer but suggests significant circadian dysregulation, often paired with insomnia and shift-work history.

High cortisol patterns

The drivers behind elevated cortisol presentations are not interchangeable, and the recovery protocol depends on which one is dominant.

Stimulant load drives a particular pattern: sharply elevated CAR, elevated midday cortisol, and frequently still-elevated evening readings if dosing extends past noon. Caffeine doses above 300 mg per day, modafinil at 100 to 200 mg, and stacked nootropic protocols can all push this picture. The ratio is often only mildly elevated because DHEA-S is preserved early in the trajectory.

Overtraining pushes a different signature. Acute and intermediate overreaching produces elevated AM cortisol with maintained or even slightly elevated DHEA-S; sustained overtraining shifts to elevated cortisol with falling DHEA-S, then eventually to suppressed cortisol with depleted DHEA-S in late-stage non-functional overreaching. The cortisol:DHEA-S ratio climbs throughout this trajectory before everything floors out. Chronically elevated cortisol also suppresses lymphocyte counts and shifts the neutrophil-to-lymphocyte ratio — changes that appear on a full blood count before hormone panels are routinely checked.

Sleep debt and caloric deficit operate through partly overlapping mechanisms. Chronic sleep restriction below 6 hours flattens the CAR while raising evening cortisol. Aggressive caloric deficits, particularly below maintenance by more than 25 percent for over 4 weeks, push AM cortisol up by 20 to 30 percent on average, with proportional DHEA-S declines.

Low and flat cortisol patterns

Low cortisol is more concerning than high, and harder to interpret. Three patterns dominate.

Post-burnout HPA suppression presents with low AM cortisol (often below 8 mcg/dL), flat saliva curve, depleted DHEA-S, and a markedly reduced ratio. This is not Addison's — ACTH is typically normal or low rather than elevated — but functional axis suppression after years of overdrive.

Exogenous glucocorticoid suppression (inhaled steroids, topical steroids over large body areas, oral courses) produces low cortisol with relatively preserved DHEA-S and low ACTH. Recovery follows a predictable timeline: weeks to months depending on duration and dose.

Late-stage overtraining looks similar to burnout but with a clearer training-load history and faster recovery once load is reduced.

Recovery protocol and tracking

Once you have identified the pattern, recovery tracking has its own rhythm. DHEA-S moves slowly. Expect 3 to 6 months for meaningful changes; tracking weekly is wasted effort. Cortisol shifts faster — 4 to 8 weeks of corrected sleep, reduced training load, and cleared stimulants will visibly normalise the AM serum value and the saliva curve. The ratio improves last because both numerator and denominator are moving.

Confounders to watch: any acute illness in the 48 hours before draw, alcohol within 24 hours, intense exercise within 12 hours, and any new supplement (especially adaptogens, licorice root, phosphatidylserine) started in the prior 4 weeks. All of these distort tracking. Standardise draw time, fasting status, and the prior 24-hour load if you want the trend to mean anything.

The takeaway is not that cortisol testing is useless — it is that cortisol testing without DHEA-S, ratio analysis, and ideally a saliva curve is barely testing at all. The HPA axis is a system, and systems require multiple data points sampled at known times to characterise. One number at one time point is a snapshot of weather, not a reading of climate. For how adrenal markers sit within a full biomarker panel — including metabolic, thyroid, and liver markers — see the complete guide to interpreting blood test results.

For Australians looking to access AM cortisol and DHEA-S without waiting for a GP referral, our guide to private blood testing in Australia covers which self-referral pathology providers include adrenal markers on their menus, what to expect on costs, and how to ensure draw timing is standardised correctly.