There's a particular frustration that comes from getting a cortisol test back — the number sits within the reference range, your GP nods, and you leave feeling no closer to understanding why you feel exhausted by 11am, why stress weeks reliably spike your cravings, or why a run of poor sleep seems to amplify every PCOS/PMOS symptom you have.
Cortisol is the most-discussed hormone in PCOS/PMOS communities after oestrogen and androgens. It also generates more confusion than almost any other test result. Many women who present with classic high-cortisol symptoms — central weight gain, sleep disruption, mood volatility, carb cravings — have normal or even low serum cortisol on a standard morning blood draw. That apparent contradiction isn't a contradiction at all once you understand what that test is and isn't measuring.
This post covers four things: why cortisol is central to PCOS/PMOS biology (not just a side effect of stress), what the different tests actually capture, why daily tracking data fills a gap that no single lab result can, and what to specifically ask for — from your GP and from yourself.
1. The HPA Axis and PCOS/PMOS — Why Cortisol Is Central, Not Peripheral
PCOS/PMOS is not primarily an ovarian condition. It's a neuroendocrine condition. The ovarian manifestation — the cysts, the anovulation — is downstream of systemic hormonal dysregulation that involves the hypothalamus, pituitary, adrenal glands, and ovaries simultaneously. The HPA (hypothalamic-pituitary-adrenal) axis sits at the centre of this.
The adrenal subtype: a significant minority
Approximately 20–30% of PCOS/PMOS cases have elevated DHEA-S (dehydroepiandrosterone sulphate) driven by adrenal androgen excess, not just ovarian androgen production. This matters because cortisol and DHEA-S share the same upstream precursor: pregnenolone. Via the 11β-HSD (11 beta-hydroxysteroid dehydrogenase) pathway, chronic HPA activation can shift the pregnenolone pool toward both cortisol and adrenal androgens. This is the adrenal subtype — and for these women, the androgen excess is literally rooted in the stress response system.
Cortisol-androgen crosstalk without elevated serum cortisol
Here's the mechanism that explains a lot of confusion: chronic HPA activation raises CRH (corticotropin-releasing hormone), which raises ACTH (adrenocorticotropic hormone), which drives adrenal androgen production — DHEA-S and androstenedione — even without measurably elevated serum cortisol. The adrenal glands are responding to the HPA signal, but the cortisol itself is being rapidly cleared or converted to inactive cortisone at the tissue level. The downstream consequences (elevated DHEA-S, worsening insulin resistance, androgen amplification) are present. The serum cortisol level looks normal. This is not a rare edge case — it's why a normal cortisol result does not mean the HPA axis is functioning normally.
What Atlas captures that maps to this mechanism
The daily signals that reflect HPA tone are not dramatic. They're consistent. Your stress score composite tracks the subjective load. Morning energy reflects the overnight cortisol arc — specifically whether the arc bottomed out properly overnight and rose appropriately at waking. Mood volatility across a week is a sensitive proxy for HPA dysregulation. And cravings severity on high-stress days directly maps to the cortisol → ghrelin → carb cravings cascade. None of these require a lab test to observe.
2. What the Different Cortisol Tests Actually Measure
Understanding what you're asking for — and what the result means — changes how useful any cortisol test is.
Serum cortisol (morning)
This is what your GP orders. A single blood draw, usually between 8–9am, measuring total cortisol in serum. Normal range at 6am: 140–690 nmol/L. The range is enormous. A result at 180 nmol/L and a result at 620 nmol/L are both "normal" — but they represent very different HPA states. More importantly, a single morning snapshot tells you nothing about: whether the Cortisol Awakening Response (CAR) occurred appropriately, how cortisol is behaving through the rest of the day, whether evening cortisol is declining as it should, or whether total daily production is elevated even if the morning snapshot looks average. Useful for: ruling out Addison's disease or Cushing's syndrome at the extremes. Not useful for: understanding subtle HPA dysregulation in PCOS/PMOS.
24-hour urinary free cortisol (UFC)
This collects all urine over 24 hours and measures total free cortisol production across the day. It integrates what the morning snapshot misses. However, its sensitivity for subtle HPA dysregulation is poor — it's primarily useful for confirming or ruling out Cushing's syndrome, where cortisol is dramatically elevated. In PCOS/PMOS, where HPA dysregulation is often more about the pattern than the total level, UFC adds limited clinical information. It's rarely worth pursuing unless Cushing's is genuinely suspected.
Salivary cortisol × 4 daily curve
This is the most informative test for understanding HPA function in everyday life, and it's not available on the NHS — it requires a private lab (Regenerus Diagnostics and BioLab both offer it in the UK, typically £60–120).
The protocol involves four saliva samples: at waking, 30 minutes post-waking, midday (12pm), and evening (8pm). The 30-minute post-waking sample is specifically designed to capture the Cortisol Awakening Response (CAR) — the physiological surge of 50–160% that should occur in the first 30 minutes after waking. The CAR is not just "cortisol is high in the morning" — it's a specific, evolutionarily conserved HPA response to waking that primes the immune system, metabolism, and cognitive readiness for the day.
What the four-point curve reveals that a morning serum draw cannot:
Flat CAR (minimal rise from waking to +30 minutes) = HPA blunting. Associated with chronic stress, burnout, and — importantly for PCOS/PMOS — prolonged HPA dysregulation. A flat CAR correlates with the morning energy crash pattern, poor cognitive readiness in the first hour post-waking, and impaired cortisol-driven cortisol feedback. This is the pattern most consistent with feeling tired despite sleeping.
Elevated evening cortisol (8pm sample above ~5 nmol/L salivary) = the arc isn't declining as it should. This has downstream consequences for both sleep (cortisol competes with melatonin) and androgen production (cortisol remaining elevated in the evening continues to drive adrenal DHEA-S). This directly maps to the sleep disruption + androgen amplification loop that characterises a significant subset of PCOS/PMOS.
If you do get a salivary cortisol curve, cross-reference the results with your Atlas data: the CAR quality maps to Atlas morning energy and morning mood composite; elevated 8pm cortisol maps to sleep quality rating and next-day energy floor.
DUTCH test (Dried Urine Test for Comprehensive Hormones)
The DUTCH test is the most comprehensive cortisol assessment currently available outside a research setting. It measures: free cortisol, cortisol metabolites (a-THF, b-THF, and THE — the hepatic breakdown products of cortisol), the cortisone ratio, and — depending on the version ordered — oestrogen metabolites and progesterone. Cost in the UK: approximately £200–400 privately.
The most clinically useful addition over the salivary curve is the 11β-HSD activity insight. High cortisol metabolites combined with low free cortisol indicates high cortisol production being rapidly converted to inactive cortisone at the tissue level — this is relevant to PCOS/PMOS because 11β-HSD activity is altered in adipose tissue in insulin-resistant states. In other words: the adrenal glands are producing more cortisol than the serum level suggests, but it's being cleared quickly. This explains the pattern of high-cortisol symptoms (central adiposity, sleep disruption, cravings) with normal-range serum cortisol.
For most women in the initial investigation phase, the 4-point salivary curve is more actionable and considerably cheaper. The DUTCH adds value if you've already done the salivary curve and want to understand cortisol production vs clearance specifically.
3. Why Tracking Data Fills the Gap Labs Leave
A lab result captures one moment. Your HPA axis operates across a life. The CAR alone varies 30–50% day-to-day depending on sleep quality, perceived stress the previous evening, and immune challenge. A single salivary test — even a four-point curve — is one day's data. It may or may not be representative.
This is not a reason to avoid testing. It's a reason to understand what testing can and cannot tell you.
Atlas captures daily: morning energy (a sensitive proxy for CAR quality), mood composite (HPA tone through the day), stress score, sleep quality (the input that most reliably predicts next-morning CAR), and cravings severity (the cortisol → ghrelin → carb cravings cascade that runs in the background of every high-stress day). These aren't approximations of a lab test — they're a different kind of data. Where a lab test gives you precision at one point in time, tracking gives you pattern across time.
The practical question is never "is my cortisol high?" — it's "on which days and under which conditions does my HPA output worsen my PCOS/PMOS symptoms?" That question cannot be answered by a blood draw. It can be answered by 28 days of consistent logging.
A pattern that Atlas can surface: on days following fewer than six hours of sleep, morning energy drops by an average of 1.8 points AND cravings severity increases by 2.1 points — consistent across the last 28 days. That's more actionable than a cortisol number in a reference range. It tells you the specific trigger, the specific downstream effect, and the direction of an intervention that's worth trying.
The sleep-cortisol connection runs in both directions: poor sleep blunts the CAR; an already-dysregulated HPA axis elevates evening cortisol and makes it harder to sleep. Tracking both sleep quality and morning energy allows you to see which direction the loop is running in your particular biology.
4. DHEA-S and Adrenal Androgen Testing — The Cortisol Partner Labs
If you're investigating adrenal involvement in your PCOS/PMOS, cortisol alone doesn't complete the picture. The adrenal panel should include:
DHEA-S — the primary marker of adrenal androgen excess. Elevated DHEA-S (above approximately 10 μmol/L in premenopausal women, though lab reference ranges vary) suggests adrenal contribution to the androgen excess that characterises PCOS/PMOS.
Androstenedione — the direct adrenal androgen precursor to both testosterone and oestrogen. Elevated androstenedione in the context of elevated DHEA-S confirms adrenal origin.
17-OHP (17-hydroxyprogesterone) — this marker is essential and often overlooked. Late-onset congenital adrenal hyperplasia (CAH) mimics PCOS/PMOS in approximately 2–4% of cases. 17-OHP should be drawn fasting, in the early follicular phase if possible. If baseline 17-OHP is elevated above approximately 6 nmol/L, the next step is a synacthen stimulation test — 250 micrograms synthetic ACTH given IV or IM, with cortisol and 17-OHP measured at 0 and 60 minutes. This is the definitive test for late-onset CAH. It matters clinically because late-onset CAH is treated differently to PCOS/PMOS (with low-dose glucocorticoids, not the standard PCOS/PMOS protocol).
ACTH stimulation test — more broadly, the synacthen test measures adrenal reserve. A flat cortisol response to synthetic ACTH suggests adrenal insufficiency; an exaggerated 17-OHP response confirms CAH. If your GP suspects adrenal involvement but isn't sure how to proceed, this is the test that provides the definitive picture.
In Atlas, the signals that act as androgen proxy markers — skin clarity (acne severity), hair shedding frequency, mood volatility on high-stress weeks — are worth tracking specifically in relation to your stress score. If these markers cluster with high-stress periods rather than with cycle phase, that's the pattern consistent with adrenal amplification: stress driving cortisol driving DHEA-S driving androgen symptoms. Tracking this pattern over 8–12 weeks gives you a directional read on whether adrenal involvement is likely before you have lab confirmation.
The insulin resistance connection matters here too: insulin resistance amplifies adrenal androgen sensitivity, meaning the cortisol-androgen crosstalk hits harder in insulin-resistant PCOS/PMOS. Improving insulin sensitivity — through dietary changes, inositol, or metformin — reduces the tissue-level amplification even without directly suppressing cortisol.
What to Ask Your GP
The adrenal panel request:
"I have PCOS/PMOS and I'm concerned there may be an adrenal component to my androgen excess. Can I have a morning cortisol (8–9am, fasting), DHEA-S, androstenedione, and 17-OHP? I'd like to rule out late-onset congenital adrenal hyperplasia."
This framing does three things: it provides the clinical rationale (ruling out late-onset CAH is a legitimate investigation), it specifies the correct timing (8–9am fasting), and it names the tests clearly so nothing gets lost in the request.
If baseline 17-OHP is above 6 nmol/L, follow up with: "Given the elevated 17-OHP, can I be referred for a synacthen stimulation test?"
What to consider privately:
The 4-point salivary cortisol curve (typically £60–120 from Regenerus or BioLab) is the most actionable next step for most women investigating HPA function beyond the morning serum draw. It answers the questions a single blood draw cannot: whether the CAR is occurring, and whether evening cortisol is appropriately declining. This is more useful as a first private investigation than a DUTCH test — reserve the DUTCH for after you've seen the pattern, if you want to understand production versus clearance specifically.
How to read your cortisol result:
A morning serum cortisol within the reference range (140–690 nmol/L) does not mean the HPA axis is functioning normally. It means total cortisol at that one moment was within a broad population range. The arc — how cortisol rises at waking, behaves through the day, and declines by evening — matters more than the level at any single point. Normal serum cortisol and significant HPA dysregulation coexist frequently.
Practical: What to Track in Atlas in Parallel
The four Atlas signals that together build a picture of HPA tone better than any single morning blood draw:
Stress score — the daily HPA input. Track at consistent times. Look for whether high-stress days consistently precede worse morning energy and higher cravings the following day.
Morning energy — the primary downstream output of the overnight cortisol arc and CAR. A chronic morning energy floor of 2–3/5 despite adequate sleep duration is a consistent signal of HPA dysregulation.
Sleep quality — the critical input variable. The cortisol arc is highly sleep-dependent. Poor sleep degrades the CAR; a dysregulated CAR worsens subsequent sleep. See the circadian reset post for the specific mechanisms.
Cravings severity — the most sensitive daily signal for cortisol-ghrelin-androgen interactions. Specifically, track whether cravings spike on high-stress days or on the day following poor sleep. If they do, that's the cortisol pathway active in your biology.
Additionally: alcohol disrupts the cortisol arc even at 1–2 drinks — it delays the overnight arc by 1–2 hours, meaning the 8am waking response starts from an already-elevated base. If you drink periodically, track alcohol as a variable and observe whether it predicts the morning energy and cravings pattern the next day.
An 8-Week Structured Experiment
Weeks 1–2: Baseline
No intervention changes. Log daily: stress score, morning energy (1–5), sleep quality (1–5), and cravings severity (1–5). This is your personal reference point. Do not skip this phase — it's what makes the rest of the data meaningful.
Week 3: Get your labs
Book a morning cortisol + DHEA-S draw, fasting, 8–9am. Ideally add androstenedione and 17-OHP to the same draw. Note your result in your Atlas tracking notes for context when you review the pattern.
Week 3 onwards: Introduce one intervention
Choose one of: introducing ashwagandha (KSM-66, 300–600mg daily — the best-evidenced HPA-modulating supplement for HPA blunting), a consistent sleep and wake schedule (same times seven days a week, including weekends), or reducing evening screen exposure after 9pm. One variable. One change.
Week 8: Review
Look at your Atlas weekly averages versus the weeks 1–2 baseline. Primary signals to check: has your average morning energy floor shifted? Has the correlation between high-stress days and high-cravings days weakened? Has sleep quality average improved? Secondary: has mood volatility on high-stress weeks reduced?
This is not a clinical trial. It's a personal experiment with real data — far more informative than a single cortisol result, and far more actionable than general advice to "reduce stress."
The Honest Framing
A morning cortisol test is worth getting — not because it will definitively answer the question, but because it's the accessible starting point and it screens out the extremes (Addison's, Cushing's). The adrenal panel (adding DHEA-S, androstenedione, 17-OHP) is the genuinely useful extension of that test for PCOS/PMOS. The 4-point salivary curve is the most informative private investment for HPA pattern. And daily tracking is what connects all of it to your actual lived experience across time.
Normal serum cortisol is not reassuring. The arc matters more than the level. And the pattern your tracking data builds across 28 days tells you something that no single lab result — however comprehensive — can tell you.
Atlas tracks the daily signals that map to your cortisol arc — stress, energy, cravings, sleep. Start building your pattern today at momentum.madethis.app.