Overtraining syndrome and cortisol testing: what the sports-medicine evidence actually shows

Cortisol is the body’s primary glucocorticoid hormone, released by the adrenal glands under control of the hypothalamic-pituitary-adrenal (HPA) axis. It follows a normal diurnal rhythm, highest shortly after waking and falling across the day, and it rises predictably in response to both physical and psychological stress, including a single hard training session. That acute, exercise-linked rise is well documented and is part of why cortisol became the go-to marker for “how hard was that, and how well am I recovering from it.”^[3]

The leap that is not well supported is treating a single cortisol reading, or a ratio built from it, as a diagnosis of overtraining. That gap between a real physiological signal and the product built to sell its measurement runs through this entire topic.

Sports medicine distinguishes several stages along the same spectrum. Successful training requires overload, but overload without enough recovery is where the trouble starts. Functional overreaching is a short-term performance dip from a deliberately hard training block that resolves within days to weeks and leads to improved performance once recovery happens; it is a normal, even useful, part of periodized training. Non-functional overreaching (NFOR) occurs when training and recovery fall further out of balance, and the athlete needs longer to bounce back. Overtraining syndrome (OTS) sits at the far end: a state the 2013 joint consensus statement from the European College of Sport Science and the American College of Sports Medicine describes as “prolonged maladaptation” of the athlete and of several biological, neurochemical, and hormonal regulation mechanisms at once.^[1]

The consensus statement is also candid about where the evidence runs out. It notes that OTS symptoms, such as fatigue, performance decline, and mood disturbance, are generally thought to be more severe than those of NFOR, but adds plainly: “there is no scientific evidence to either confirm or refute this suggestion.” Telling OTS apart from NFOR is, in the statement’s own words, “very difficult” and depends on how a person responds over time and on ruling out other causes, not on a single test or symptom.^[1]

This is the core finding worth sitting with: a panel of sports-science and sports-medicine experts from two major professional bodies specifically set out to evaluate the markers used to detect overtraining syndrome, hormonal, performance-based, psychological, biochemical, and immune, and concluded that “currently, several markers… are used, but none of them meet all the criteria to make their use generally accepted.”^[1] Cortisol is one of the hormonal markers in that mix, not an exception to it.

A 2017 systematic review by Cadegiani and Kater pooled 38 studies that measured hormones in athletes with overtraining syndrome, functional overreaching, or non-functional overreaching. The headline finding: basal levels of hormones, measured at rest, were mostly normal in these athletes compared with healthy, non-overtrained athletes. Cortisol specifically, along with plasma catecholamines, showed conflicting results across studies, while most other hormones responded normally. The one signal that looked more consistent came from stimulation tests under maximal exercise conditions, where blunted growth hormone and ACTH responses showed up more reliably in overtrained athletes.^[2]

That matters directly for the cortisol:testosterone ratio marketed by some recovery-testing services and wearables, on the theory that a shift toward cortisol and away from testosterone reflects tissue breakdown outpacing repair. If basal testosterone, like basal cortisol, is typically normal in overtrained athletes, a one-off ratio built from both numbers inherits the same weakness. Day-to-day variation, sleep, recent training load, and the time of day a sample is drawn all move both hormones independently of whether someone is actually overtrained.

Saliva sampling is the non-invasive method behind most consumer-facing cortisol and recovery tests, since it avoids a blood draw. A 2023 systematic review focused specifically on salivary biomarkers during the post-exercise recovery period, pooling 14 studies out of 213 identified. Cortisol and testosterone, along with their ratio, were the most frequently studied markers. The review found that salivary markers clearly respond to exercise, but that their usefulness for gauging recovery status specifically “is still inconclusive,” with normalization timelines varying by exercise intensity, duration, and testing method across studies.^[3]

None of this means saliva testing is a scam; it is a legitimate, active area of sports-science research. It means the research has not yet produced a validated threshold or protocol that a commercial product can hand you as a reliable, individual readiness score.

It is worth being precise about what overtraining syndrome is not. A separate, much broader wellness-industry label, “adrenal fatigue,” is often applied to general tiredness and burnout in the public, typically “diagnosed” through salivary or urinary cortisol panels sold directly to consumers. A 2016 systematic review by the same research group, covering 58 studies and more than 3,470 articles screened, examined whether cortisol testing substantiates adrenal fatigue as a condition. It found an almost systematic pattern of conflicting results regardless of which test was used or how rigorous it was, and concluded plainly: “this systematic review proves that there is no substantiation that ‘adrenal fatigue’ is an actual medical condition. Therefore, adrenal fatigue is still a myth.”^[4]

Overtraining syndrome, by contrast, is documented in a formal joint consensus statement from two major sports-science and sports-medicine bodies, tied to a specific, verifiable history of training overload, and taken seriously as a real, if genuinely hard to pin down, clinical picture. The two labels get conflated constantly in marketing copy. They are not the same claim, and they do not carry the same evidence behind them.

The consensus statement frames prevention and treatment around restoring the balance between training load and recovery, not chasing a single number back to normal. That includes planned reductions in training volume or intensity, attention to sleep, adequate carbohydrate and overall caloric intake, and ruling out other contributors such as iron or magnesium deficiency, infection, or an underlying illness. Confirmed overtraining syndrome can require a complete rest period lasting weeks to months, considerably longer than the days-to-weeks recovery that resolves ordinary functional overreaching.^[1]

We have made a version of this same point elsewhere on this site about a different kind of readiness decision. In Return to sport: how to know a tendon is ready, pain relief alone is not a reliable signal that a tendon has actually healed; restored strength symmetry, demonstrated load tolerance, and time all matter more than any single check-in. The same logic applies here: tracking training load, performance, sleep, and mood over weeks tells you far more than one cortisol reading ever will.

Overtraining syndrome is real, and the mismatch between hard training and insufficient recovery that causes it is worth taking seriously. What is not supported by the evidence is the idea that a single cortisol test, or a cortisol:testosterone ratio, can diagnose it. The expert panel that actually reviewed the available markers concluded none of them meets the bar on its own, and the hormone-specific research since then backs that up: basal cortisol is often unremarkable in overtrained athletes, and salivary testing remains an active research question rather than a settled, individually predictive tool. That is a genuinely different, weaker claim than “adrenal fatigue,” which the cortisol-testing literature does not support at all. See how we grade the strength of evidence across topics in our editorial and evidence standards, and browse our Hormone & Vitality hub for related coverage.

None of this replaces an individual conversation with a clinician who knows your training history, symptoms, and lab values. It reflects what the consensus guidance and the completed systematic reviews actually show as of this writing.