The shape of the 24-hour cortisol curve, the awakening response, and what chronic stress does to both.
What the evidence says
- Cortisol follows a circadian rhythm: it begins rising in the early morning before waking, peaks in the morning hours around awakening, then declines across the day in two phases to a nocturnal nadir[1]
- In healthy people the majority of the day’s cortisol is secreted in the several hours surrounding morning awakening — the morning is not just the high point of the curve, it is where most of the output happens[1]
- The rhythm is endogenous — driven by the body clock, not merely by the act of sleeping and waking. In forced-desynchrony laboratory protocols, cortisol rose across the circadian phases corresponding to habitual sleep and peaked at the phase corresponding to shortly after habitual wake time (participants’ mean habitual wake time was 6:36 a.m.; the study gives no clock time for the peak itself)[2]
- The cortisol awakening response (CAR) is a distinct phenomenon: the rapid rise across the first 30–45 minutes after waking, reported as roughly a 50%–156% increase in free salivary cortisol[1]
- How broadly that has been observed, which sorts this file’s claims by weight: the CAR emerges as a particularly large cortisol burst across samples from 18,698 unselected individuals — a different order of evidence from the sleep-laboratory work two bullets down, which the review itself says requires replication[1]
- Critically, the CAR is not a name for the whole daily curve — it is an awakening-entrained ultradian pulse superimposed on the diurnal rhythm, and is largely unrelated to general diurnal cortisol measures[1]
- The evidence that the CAR is genuinely separate rather than a continuation of the pre-awakening rise comes from sleep-laboratory work that the review itself says requires replication — the distinction is well-supported but not settled[1]
- Morning light exposure after waking has been linked to a larger CAR, most strongly for short-wavelength (blue) light — but a significant CAR still occurs in complete darkness, which the review takes as ”suggesting that the CAR is not just a response to light and that light and awakening-induced influences are complementary” (corrected 2026-09-10 from ”so light modulates the response rather than causing it”, the library’s reading)[1]
- What the CAR is for is not settled. Energy mobilisation is described as a preeminent effect of cortisol generally, but the review calls direct investigation of the CAR’s own metabolic function scarce (one study failed to show that lower awakening glucose predicts a greater CAR), and its “mobilise resources for the day’s demands / counterregulate the previous day’s adverse emotional experiences” account is offered explicitly as a model rather than an established finding[1]
- The CAR is itself circadian-gated: it is largest when waking falls at a circadian phase of roughly 3:40–3:45 a.m. — about three hours before habitual wake time — and there is effectively no CAR when someone wakes at a circadian phase corresponding to the afternoon or evening. Waking alone does not produce one[2]
- That gating is a group-level finding, and should be stated as a tendency rather than a law: individually significant rhythms appeared in only 60% and 71% of participants across the two protocols, the effect size differed more than twofold between them, and the combined sample was 35 adults[2]
- Chronic stress blunts the CAR, and chronic adversity and negative health conditions ”appear to reduce the dynamic range of circadian cortisol secretion” — the peak-to-trough difference shrinks, which is the “flattened curve” pattern (corrected 2026-09-10: appear to restored)[1]
- The blunting finding is comparatively well-established: the review attributes it to high-quality research, including students under long-term examination stress, and notes flattened CAR profiles across multiple adverse conditions[1]
- The diurnal cortisol slope — the degree of change from morning to evening across the waking day — is sensitive to emotional and psychosocial stress; both adverse experience and worse health track a flattened slope[3]
- What “flattened” physically consists of is not one pattern: flatter slopes contain elements of both hypo- and hyper-cortisolism, meaning a low morning and/or a high evening. So an elevated evening is one way the curve flattens — but the evidence does not support the common claim that chronic stress simply keeps cortisol high all night[3]
- Scale of the health link: meta-analysis of 179 associations from 80 studies — 36,823 participants, or 26,167 unique individuals once sample overlap is accounted for — found flatter slopes significantly associated with poorer health overall (r=.147) and in 10 of 12 outcome subtypes, largest for immune/inflammation (r=.288)[3]
- The two subtypes that did not reach significance were cardiovascular disease and anxiety. Depression symptoms and diagnoses did (r=.106)[3]
- A flatter slope is significantly associated with obesity, BMI and adiposity (r=.093, p=.002, from 17 findings) — real, but the small end of an already-small range. And it is a different claim from “stress raises cortisol and cortisol stores fat”: flattening can equally consist of a low morning[3]
- That link should be stated as an association, not a mechanism: 91.1% of the underlying findings are cross-sectional with cortisol and outcome measured concurrently, effect sizes are small on average, and the authors explicitly leave open whether flattening is a marker or a cause — while still calling it a “strong candidate mechanism” worth pursuing. Both halves belong in any summary of their position[3]
- The authors’ own instruction on how hard to lean on these numbers: average effect sizes “should not be regarded as exact,” but as indicators of general patterns and of the relative strength of associations across outcomes[3]