Cortisol test: what a cortisol measurement does and does not say about recovery

Cortisol is a hormone made by your adrenal glands. It releases energy, helps keep your blood sugar steady, and follows a strong daily rhythm: high in the early morning, low around midnight. It also rises during a hard training session, and that is supposed to happen. That rise is part of the load you are training for in the first place.

Measuring cortisol is a real, recognised test. Doctors use it to detect two rare conditions: Cushing's syndrome (too much cortisol) and adrenal insufficiency, or Addison's disease (too little cortisol). These are serious illnesses and the test is essential for them.

But look at what that test actually involves. The Endocrine Society guideline prescribes: two late-night saliva samples, two 24-hour urine collections, or a dexamethasone suppression test where you take a tablet the night before. None of these is a blood draw you can simply walk in for. For Addison's disease, a morning blood draw between 06:00 and 10:00 is a legitimate first step, but only alongside ACTH, and only when a doctor already suspects the disease. Those same guidelines explicitly recommend against a random serum cortisol.

This is the core problem, and it is arithmetic, not opinion.

Cortisol fluctuates enormously, even in healthy people doing nothing unusual. Clinical chemistry calls this the reference change value: how much a value must change before you can say with 95% confidence that something genuinely changed. For cortisol in blood, that is around 47%. For salivary cortisol it runs, depending on the time of day, from 96 to 245%.

Put plainly: your cortisol has to roughly double before anyone can establish that anything changed at all. Everything below that is noise. We could draw your blood twice, an hour apart, and get two numbers tens of percent apart, both entirely normal, and both meaningless.

That is exactly what makes cortisol useless as a recovery metric. The idea is appealing: draw every Monday morning, plot the curve next to your HRV, see when you need a deload. But the difference between a good and a bad training week is small, and the day-to-day noise in your cortisol is large. The noise is bigger than the signal you are looking for. You would not be measuring your recovery, you would be measuring the day.

On top of that, the things that shift your cortisol are precisely the things you bring with you:

• the training itself: a hard session lifts your cortisol temporarily, so whether you draw the morning after a deadlift day or after a rest day changes your number
• sleep loss and an irregular rhythm, which overturn the daily rhythm the test leans on
• the contraceptive pill and other oestrogen-containing contraception, which raise a binding protein in your blood so your measured total cortisol goes up while nothing is actually wrong
• the needle itself: the stress of the blood draw raises your cortisol at the very moment of measurement

This is why most lifters end up here: cortisol is catabolic, so a high value means you are losing muscle.

Half of that is true. At pathologically high cortisol levels, as in Cushing's syndrome or long-term high-dose prednisone, muscle breakdown does occur. That is real and well described. But those are levels that stay extremely elevated for weeks to months. The spike you see in your blood after a hard session is something else entirely: a temporary response to load that fades within hours.

So what happens when you actually measure it? In a study of 56 young men doing twelve weeks of resistance training, the acute hormone response after a hard leg session was correlated with their gains in muscle and strength. The acute rise in growth hormone, free testosterone and IGF-1 showed no correlation with gains in lean body mass or strength. Cortisol, in the researchers' words "usually assumed to be a hormone indicative of catabolic drive", was positively correlated with the gain in lean body mass (r = 0.29) and with type II fibre cross-sectional area (r = 0.35).

Please do not read that as "cortisol builds muscle". The correlation is weak and the study is small. The honest conclusion is this: the acute hormone response to training barely predicts anything, and what is measured certainly does not point the way the myth claims. No study shows that a normal cortisol value is costing you muscle.

The testosterone-to-cortisol ratio, proposed as an overtraining marker for decades, has the same problem: you are stacking the noise of two fluctuating hormones on top of each other. The joint consensus statement of the European College of Sport Science and the American College of Sports Medicine is clear about it. Many markers are in use (hormones, performance tests, psychological tests, biochemical and immune markers), but none of them meets all the criteria to be generally accepted.

Suppose you draw anyway and a number comes back. Which way does it point?

That is the question the whole idea breaks on, because the direction of the cortisol effect is not even stable. The largest meta-analysis in this field reviewed 147 studies of the cortisol response after waking. That response is higher with job stress and general life stress, and lower with fatigue, burnout and exhaustion. The same axis, two opposite directions, depending on where in the story someone happens to be.

And the best evidence that actually argues for cortisol points somewhere else again. A 2018 Austrian study compared 40 people with burnout to 26 healthy controls and found higher salivary cortisol in the burnout group, most clearly around midday and at the daily nadir. That is the exact opposite of the popular story in which "exhausted adrenals" make less cortisol.

With that, the floor drops out from under the measurement. A single value has no interpretable direction. High could be stress, could be your commute, could be the needle, could be yesterday's session. Low could be nothing, could be the time of day, could be the exhaustion end of that same curve. You cannot distinguish "overtrained" from "stressed", and neither from "nothing is wrong", because one number simply does not carry that information.

A systematic review of the biomarker studies in burnout (31 studies) reached the same conclusion: no potential biomarkers for burnout were found. And "adrenal fatigue", the idea that chronic stress burns out your adrenal glands so they produce too little cortisol, does not exist. A 2016 systematic review searched 3,470 publications, kept 58, and found no substantiation at all; in most fatigued participants the cortisol values were simply normal. This is not an academic quibble: a study in Mayo Clinic Proceedings tested twelve over-the-counter "adrenal support" supplements and found thyroid hormone that was not on the label in all twelve, and a steroid in a quarter of them. If you are drug-tested, that is not merely pointless, it is a risk.

The need behind your search is legitimate. You want to know whether you are doing too much and whether your recovery is falling behind. Blood cortisol is simply the wrong instrument for it.

The things that are trackable share three properties: you can measure them daily, compare them against your own baseline, and they move enough to be visible through the noise.

• HRV and resting heart rate: every morning, at home, against your own rolling average. Neither is a diagnosis of overtraining, and the consensus statement above applies to them too, but you can at least see a trend. With cortisol, your value has to nearly double before anything counts as a change, which is why you never see a trend in it.
• Sleep duration and sleep regularity, plus your subjective recovery score. Simple, boring, and by far the strongest predictors of how your week will go.
• Your training load and your energy intake. Ramping up too fast and chronically under-eating explain more flat sessions than any hormone does.

If you do want blood work because your fatigue persists, this is the list Dutch primary-care diagnostics starts with. Short, and for a reason:

• Haemoglobin, and if abnormal your iron stores via ferritin: anaemia and iron deficiency are the most commonly found physical cause, and low ferritin is common in endurance athletes. It is also genuinely treatable.
• CRP: an inflammatory marker, which makes an infection or inflammatory disease less likely
• Fasting glucose: to rule out diabetes
• TSH: your thyroid, which genuinely is a common and treatable cause of fatigue

We would rather earn nothing from you than sell you a test that does not answer your question.

There genuinely are situations where measuring cortisol is the right thing to do. If you recognise yourself here, you can order the test below:

• your GP or specialist asked for a cortisol measurement. Then you are here to have the test done, not to diagnose yourself, and we are glad to help
• you already have a known adrenal or pituitary condition and you are monitoring your values
• you use or have stopped corticosteroids (prednisone, dexamethasone, including creams and inhalers), where your own cortisol production may be suppressed

If you order without a doctor's request, book your draw between 06:00 and 10:00 in the morning, fasting, without heavy exertion beforehand, and not the morning after a hard session. Outside that window your result cannot be compared against the correct reference range.

One thing holds even then: a single cortisol value is never a diagnosis. An abnormal result means your GP has to redo the work properly, with the right test.

We think you should be able to check what we base this on.

• Chida Y, Steptoe A. Cortisol awakening response and psychosocial factors: a systematic review and meta-analysis. Biol Psychol. 2009;80(3):265-78. PMID 19022335. (147 studies: the response rises with job stress and falls with fatigue, burnout and exhaustion.)
• Pilger A et al. Midday and nadir salivary cortisol appear superior to cortisol awakening response in burnout assessment and monitoring. Sci Rep. 2018;8(1):9151. PMID 29904183. (The burnout group had higher, not lower, cortisol.)
• Danhof-Pont MB et al. Biomarkers in burnout: a systematic review. J Psychosom Res. 2011;70(6):505-24. PMID 21624574. ("No potential biomarkers for burnout were found.")
• West DWD, Phillips SM. Associations of exercise-induced hormone profiles and gains in strength and hypertrophy in a large cohort after weight training. Eur J Appl Physiol. 2012;112(7):2693-702. PMID 22105707.
• Meeusen R et al. Prevention, diagnosis, and treatment of the overtraining syndrome: joint consensus statement of the ECSS and the ACSM. Med Sci Sports Exerc. 2013;45(1):186-205. PMID 23247672. (No proposed marker meets all the criteria for general use.)
• Cadegiani FA, Kater CE. Adrenal fatigue does not exist: a systematic review. BMC Endocr Disord. 2016;16(1):48. PMID 27557747.
• Akturk HK et al. Over-the-Counter "Adrenal Support" Supplements Contain Thyroid and Steroid-Based Adrenal Hormones. Mayo Clin Proc. 2018;93(3):284-90. PMID 29502560.
• Nieman LK et al. The Diagnosis of Cushing's Syndrome: An Endocrine Society Clinical Practice Guideline. J Clin Endocrinol Metab. 2008;93(5):1526-40. PMID 18334580. (Explicitly recommends against random serum cortisol.)
• Bornstein SR et al. Diagnosis and Treatment of Primary Adrenal Insufficiency: An Endocrine Society Clinical Practice Guideline. J Clin Endocrinol Metab. 2016;101(2):364-89. PMID 26760044. ("There is no evidence to support the use of random cortisol.")
• NHG/NVKC LESA Laboratoriumdiagnostiek, chapter on general blood work.