The Cortisol Awakening Response in Athletes: What Morning Cortisol Reveals About HPA Axis Health and Overtraining Risk
Table of Contents
Introduction
In clinical sports medicine, the evaluation of training-related fatigue, recovery capacity, non-functional overreaching and overtraining syndrome remains a clinical assessment based on the full picture: symptoms, performance trends, training history, recovery markers, and targeted laboratory testing. No single blood test, hormone value or isolated baseline measurement can diagnose or exclude overtraining, because the relevant changes often develop over time and require knowledge of the athlete’s previous baseline. This is why dynamic markers — the kinetics of the hormonal response rather than one static value — are of particular interest.
One of the most physiologically informative examples is the cortisol awakening response in athletes: the rise in cortisol during the first 30–45 minutes after waking. The CAR is a distinct component of the circadian cortisol profile that may be influenced by training, competition, and other stressors affecting HPA-axis activity. It should not be interpreted as a standalone diagnostic test, but as a potential longitudinal marker that may add useful context when assessed alongside symptoms, performance data, recovery metrics, and other clinical findings. Its application in athlete monitoring remains a growing area of research with increasing relevance for sports medicine practitioners.
What Is the Cortisol Awakening Response? HPA Axis Physiology Explained
To interpret the cortisol awakening response in athletes meaningfully, the underlying axis must be understood. [1] The hypothalamic-pituitary-adrenal (HPA) axis is principally controlled through corticotropin-releasing hormone (CRH) secretion from the hypothalamus. AVP (arginine vasopressin) may also act synergistically with CRH to stimulate adrenocorticotropin hormone (ACTH) synthesis and secretion from the anterior pituitary. The increased ACTH concentration then activates adrenocorticotropic receptors on the adrenal cortex to stimulate secretion of the steroid hormone cortisol.
Cortisol shows strong diurnal variation, with highest levels around the period following waking. [1] Superimposed on this diurnal pattern is the cortisol awakening response (CAR): a distinct rise in cortisol observed immediately after waking, typically peaking 30–45 min after waking. This is considered a distinct component of the circadian cortisol profile, linked to waking and to anticipated demands, rather than merely a continuation of the overnight rise.
In elite male athletes, one study reported a mean CAR∆30 of 42.0% and in male controls 27.0%; the authors reported no statistically significant cohort difference in relative CAR∆30 between the two groups, and noted that all hormonal variables showed poor short-term stability in the athlete cohort. [5] This substantial day-to-day variation in elite sport contexts limits isolated single-measurement interpretation.
In practical clinical work, however, CAR is not a routine test, at least in Finland. I rarely encounter it in ordinary clinical pathways, and most general clinics do not use or interpret CAR measurements regularly. In practice, it is more often seen in research protocols or highly specialised elite-sport settings, where a clinician familiar with athlete monitoring and HPA-axis physiology has a specific reason to request it. There is also no widely standardised clinical reference range for CAR in athletes. Interpretation is therefore based less on a single universal cut-off and more on the athlete’s own baseline, longitudinal change, sampling quality, and cautious comparison with patterns reported in the research literature. For that reason, I see CAR as a specialised monitoring tool rather than a standard diagnostic investigation. Its interpretation requires careful sampling, longitudinal context, and a clinician who understands the limits of what this marker can and cannot show.
Cortisol Awakening Response in Athletes: Training Load and Stress Monitoring
Interest in the cortisol awakening response in athletes as a monitoring tool has grown substantially over the past decade. [5] In different sport and exercise settings, the CAR has shown promise as a tool for assessing stress, recovery, and fatigue, but is not yet considered a reliable clinical tool. One possible reason is that athletes, especially those involved in elite level sports, are exposed to a myriad of overlapping stressors, which can affect the stability of the CAR and any resultant trends or predictions arising from it.
A systematic review by Anderson and Wideman examined 10,292 articles and included 32 studies in the final analysis. [1] The available literature suggests a threshold of exercise may be required to alter the HPA axis and affect CAR. Moreover, CAR may represent a combination of previous exercise load and upcoming stress, making current interpretation of field-based observational research challenging. Notably, no studies in that review investigated the effects of laboratory-controlled exercise on CAR, and variable effects were observed across field studies.
The relationship between training load and the cortisol awakening response has been directly investigated in recreational endurance athletes (n = 15). [4] CAR (r² = .352, P = .025) and CAR% (r² = .373, P = .012) both showed a significant negative relationship with training load — that is, higher training load was associated with a lower CAR. The authors concluded that CAR is affected by regular exercise training loads in recreational athletes. This negative relationship supports the idea that training load can influence CAR, but should not be interpreted alone as proof of HPA-axis suppression.
In my own view, the main reason CAR is interesting is not that it would be a ready-made diagnostic test for overtraining. It is not. What makes it interesting is that it may be one of the closest practical ways we have to observe a natural cortisol response without using a pharmacological challenge.
In ordinary clinical endocrinology, when we want to test the HPA axis more directly, we often move toward specific tests such as the dexamethasone suppression test. That type of testing has its place, but it requires a clear clinical indication and is designed for specific endocrine questions, not routine athlete monitoring. It also involves medication exposure, which I do not see as a neutral step in an otherwise healthy athlete unless there is a proper reason for it.
This is where CAR becomes conceptually attractive to me. Instead of forcing the axis pharmacologically, it uses the body’s own morning transition from sleep to wakefulness as the stimulus. If the sampling protocol is followed carefully, the test can be repeated across several mornings without medication, injections, or laboratory attendance at the exact moment of waking. That makes it better suited to longitudinal observation than a single static cortisol value.
I would still be very cautious in how I use or interpret it. CAR cannot diagnose overtraining on its own, and it cannot exclude it either. It does not replace clinical judgement, training history, symptoms, performance trends, HRV, blood work, or the broader assessment of recovery. But in an athlete where the clinical question is specifically about stress physiology, recovery capacity, and possible HPA-axis dysregulation, I see CAR as a potentially useful additional layer of information.
Blunted Cortisol Awakening Response in Athletes: Competition Stress and Maladaptive Patterns
The cortisol awakening response in athletes shows one of its most diagnostically instructive patterns under competition stress. MacDonald and Wetherell studied elite rowers (N = 8) across two training and two competition weekends, measuring salivary cortisol immediately upon waking and 30 minutes post-waking alongside self-reported competitive anxiety. [2] Self-reported cognitive and somatic anxiety levels were significantly greater during the competition phase compared with training. Additionally, levels of cognitive anxiety were greater on the day of competition compared with the preparation day. CAR magnitude was significantly reduced during the competition phase compared with training; however, there were no differences between preparation and event days.
The authors’ interpretation was that this represents a maladaptive response. [2] Reduced or blunted CARs are typically observed in chronically stressed populations and are characteristic of burnout and fatigue. While an increased CAR during competition may represent an adaptive response to challenge, blunted CARs and the concomitant increases in competitive anxiety observed here indicate maladaptive responding during a period where maximised functioning is critical.
A blunted cortisol awakening response in athletes around competition, especially with elevated anxiety, may indicate maladaptive stress responding — though it is important to note that this study involved only eight athletes and caution is warranted in generalising these findings. This pattern is consistent with the literature on cortisol and overtraining and the testosterone-to-cortisol ratio as markers of accumulated physiological stress.
Clinically, this overlap matters because reduced or blunted CARs have been described in chronically stressed populations and have been associated with fatigue, exhaustion, and burnout. [2] This means that a blunted CAR in an athlete cannot automatically be attributed to training load alone.
In real clinical work, athletes with suspected non-functional overreaching or overtraining often present with symptoms that overlap with burnout-type exhaustion, sleep disturbance, low mood, reduced motivation, and general stress-related fatigue. These are not always cleanly separated categories. Competitive athletes may be exposed at the same time to high training load, psychological pressure, travel, poor sleep, life stress, and pre-competition anxiety. From my perspective, this makes CAR best understood as a stress-physiology marker rather than a sport-specific overtraining marker.
A low or blunted CAR may support the impression that the athlete’s stress-response system is not behaving normally, but it does not tell us why. The cause may be excessive training load, insufficient recovery, burnout-type stress, sleep disruption, psychological strain, or a combination of these. That is why CAR should always be interpreted alongside the clinical interview, training history, sleep assessment, mood symptoms, performance trajectory, and broader recovery markers.
Cortisol Awakening Response, Athletes, and Overtraining Syndrome: EROS-CAR Evidence
The most direct evidence linking the cortisol awakening response in athletes to clinical overtraining syndrome comes from the EROS-CAR study, a secondary analysis of the Endocrine and Metabolic Responses on Overtraining (EROS) dataset conducted exclusively in male athletes. Anderson, Wideman, Cadegiani and Kater compared salivary cortisol at waking (S1) and 30 minutes post-waking (S2) across three groups: athletes diagnosed with OTS, healthy athletes, and sedentary controls. [3] The models demonstrated significant time-by-group interaction for OTS for the 2 cortisol concentrations collected during the awakening period (β = −9.33, P < .001), but not for the diurnal cortisol slope (β = 0.02, P = .80). These results suggest the CAR may be associated with OTS and should be considered within a panel of biomarkers. The authors note that further research is necessary to determine whether alterations in the CAR may precede the diagnosis of OTS.
The group effect was detected in the awakening samples rather than the modelled diurnal slope, which has a practical implication: a two-sample awakening protocol may capture information about HPA-axis dynamics not represented by some other cortisol summaries. A standard mid-morning blood draw does not constitute a cortisol awakening response assessment.
Recovery data from the EROS-LONGITUDINAL study subsequently reported that cortisol awakening response increased (P = .001) among athletes recovering from OTS, alongside a multi-marker improvement profile. [6] This supports the possibility that CAR may reflect aspects of recovery in OTS, though the complexity of OTS remission means this should not be taken as standalone confirmation of recovery.
This connects to parallel monitoring approaches discussed in resting heart rate changes in overtrained athletes and HRV and blood work, where autonomic dysregulation provides a concurrent window into athlete recovery status.
In practical patient care, I would not generally recommend a series of repeated cortisol measurements for diagnosing overtraining syndrome or non-functional overreaching. This is not established clinical standard of care, and at this stage it remains closer to experimental or research-oriented monitoring than routine diagnostic practice. For most patients, repeated cortisol testing is unlikely to be worth the cost unless there is a very specific specialist indication.
My usual approach would be more conservative. I would first assess the clinical picture carefully and use basic laboratory testing mainly to exclude other medical causes of fatigue and underperformance. This may include a full blood count, iron status, thyroid function, and selected vitamin markers such as B12 and vitamin D when clinically relevant. I would not routinely order broad hormonal panels. Testosterone testing, for example, is reasonable only if there is a genuine clinical suspicion of hypogonadism based on symptoms, history, and examination.
In other words, laboratory testing can help rule out anaemia, iron deficiency, thyroid disease, vitamin deficiency, inflammatory illness, or other medical explanations, but it does not replace the clinical diagnosis. Overtraining syndrome and non-functional overreaching remain diagnoses based on the athlete’s symptoms, training history, performance decline, recovery pattern, exclusion of alternative causes, and the clinician’s overall judgement. CAR may be intellectually and physiologically interesting, but in routine patient care it should not be sold as a necessary or definitive test for overtraining.
Conclusion
The cortisol awakening response in athletes is a physiologically interesting marker because it reflects a natural HPA-axis response during the transition from sleep to wakefulness. It may offer useful insight into stress physiology, recovery capacity, and longitudinal changes in hormonal responsiveness, especially in research settings or highly specialised elite-sport environments. However, CAR should not be presented as a routine clinical test or a shortcut to diagnosing overtraining.
At the end of the day, overtraining syndrome, non-functional overreaching, and functional overreaching remain clinical diagnoses. They are based on the athlete’s symptoms, training history, performance decline, recovery pattern, psychological load, sleep quality, and exclusion of other medical causes. Laboratory testing can support this process, mainly by ruling out conditions such as anaemia, iron deficiency, thyroid disease, vitamin deficiency, inflammatory illness, or true endocrine disease. But no single cortisol value, CAR measurement, hormone panel, or biomarker profile can diagnose or exclude overtraining on its own.
For this reason, I would not generally recommend repeated cortisol testing as part of routine patient care for suspected overtraining. In most cases, it is not established standard practice, and it is unlikely to be worth the cost unless there is a specific specialist indication. CAR may have value as a repeatable physiological response marker when used carefully and longitudinally, but its interpretation requires strict sampling, knowledge of the athlete’s baseline, and an understanding of its limitations.
In my view, the real clinical value of CAR is not that it gives a final answer. It may help add one more layer of context when the question is specifically about stress physiology and HPA-axis responsiveness. But the diagnosis still comes from the full clinical picture — not from the test itself.
References
[1] https://pmc.ncbi.nlm.nih.gov/articles/PMC5635140/
[2] https://doi.org/10.3389/fpsyg.2019.01684
[3] https://doi.org/10.1123/ijspp.2020-0205
[4] https://doi.org/10.1123/ijspp.2017-0740
