Major depression is a highly prevalent mood disorder, affecting about 350 million people worldwide (1). It is more prevalent in women than men and has a huge impact on general health of the patients (1, 2).

Major depressive disorder has been closely associated with deregulations of the hypothalamic-pituitary-adrenal (HPA) axis, both at rest and in response to stress (3–5). Some studies report changes in cortisol response that occur soon after awakening (6). Most often reported, cortisol awakening response is increased in patients with major depression, suggesting hyperactivity of the HPA axis (7). However, there is increasing evidence of hypocortisolism in patients with depression, which has been interpreted as an indication of HPA axis fatigueness in response to recurrent depressive episodes (8, 9). Such discrepancies can be attributed to a number of factors including subtypes of depression, depression severity, sex, duration of illness and socioeconomic status (10–13).

Cortisol assessments have also served as an important biomarker of treatment response. For instance, patients with depression who responded to an 8-week treatment with fluoxetine, a selective serotonin reuptake inhibitor, presented decreased levels of cortisol (14).

Most often, patients are treated with antidepressant medications that at present take about 2 weeks for the beginning of their therapeutic effects (15–18). Recently, however, psychedelics have been emerging as a promising fast-acting antidepressant (19–23). A recent open label trial in treatment-resistant depression observed a reduction of up to 87% in depression severity, already at 24 h after a single dosing session with ayahuasca (23, 24). Ayahuasca was originally used for medicinal purposes by indigenous populations from Brazil, Ecuador, Peru, and Colombia, and later its ritualistic use became more popular by its presence in ceremonies of different syncretic churches in Brazil, which is currently spreading to other parts of the world (25).

Ayahuasca is a decoction of two plants: Psychotria viridis and Banisteriopses caapi (26). P. viridis contains the psychedelic tryptamine N,N-dimethyltryptamine (N,N-DMT), whose action is mediated by serotonin (5-HT2A) and sigma-1 receptors (27–30). B. caapi contains β-carbolines (harmaline, harmine and tetrahydroharmine), which work as indirect monoaminergic agonist due to the inhibition of monoamine oxidase isoenzyme (MAO) (31, 32). Regular ayahuasca users in religious contexts have shown low level of psychopathologies (33–35), low scores on the state scales related to panic and hopelessness (36), as well as good performances in cognitive neuropsychological tests (37, 38). Moreover, this brew does not exhibit dose tolerance, i.e., the decrease of the effect of a drug or medication by excessive or frequent exposure of the patient to its active principle, and is not addictive (39–41).

Considering that the main neurobiological actions of ayahuasca are strongly related to key physiological systems altered in major depression, i.e., was observed that ayahuasca increased cortisol levels 2 h after its ingestion in healthy subjects (42) and taking into account the low incidence of mental disorders in regular ayahuasca users in religious context and previous results from open label trial (23), we recently conducted a randomized controlled trial (RCT) with ayahuasca in patients with treatment-resistant depression. Our results suggest significant and rapid reduction in depressive symptom after a single ayahuasca session, when compared to placebo (43). Herein, we explored the effects of ayahuasca on the awakening salivary cortisol response and plasma cortisol, in patients with treatment-resistant depression and in healthy individuals.

Our hypotheses are that patients and controls will show different levels of plasma cortisol and awakening salivary cortisol response, at baseline. Furthermore, we expect that the cortisol levels in the group of patients will be correlated with the severity and/or duration of disease. Moreover, we expect that ayahuasca, but not placebo, will increase cortisol levels acutely, after the dosing session and after 48 h of its ingestion, in patients and controls. Changes in cortisol levels will be correlated with the observed improvement in the symptoms of depression.

This is a randomized double-blinded placebo-controlled trial using a parallel arm design. Patients were referred from psychiatric units of the Onofre Lopes University Hospital (HUOL), in Natal/RN, Brazil, and through media and internet advertisements. All procedures took place at the HUOL. The authors assert that all procedures contributing to this work comply with the ethical standards of the relevant national and institutional committees on human experimentation and with the Helsinki Declaration of 1975, as revised in 2008. The study was approved by the Research Ethics Committee of the University Hospital (# 579.479), and all subjects provided written informed consent prior to participation. This study is registered at http://clinicaltrials.gov (NCT02914769).

Socio-demographic characteristics of healthy controls (C), and patients with major depression (MD) are summarized in Table 1. All volunteers (n = 71; MD = 28, C = 43) are Brazilian, adults [MD = 41.54 ± 11.55, C = 31.21 ± 9.87 years, t₍₆₉₎ = 4.03 p < 0.0001]. From January 2014 to June 2016, 218 patients were screened for eligibility, and 35 met criteria for the trial. On average, patients presented 11.03 ± 9.70 years of depressive symptoms and met criteria for moderate-to-severe depression (HAM-D = 21.83 ± 5.35). They were treated previously with 3.86 ± 1.66 different types of antidepressants and two patients used electroconvulsive therapy as treatment. The majority of patients presented comorbidity, such as personality disorder (76%) and anxiety disorder (31%).

Patients showed significant lower socioeconomic status backgrounds than healthy volunteers: large part of the MD group was from low socioeconomic status, unemployed [MD = 54%, C = 12%, X²(1) = 4.74, p < 0.0001], earning up to 5 minimum wages [MD = 87%, C = 71%, X²(3) = 14.03, p = 0.003] and with up to 8 years of formal education [MD = 39%, C = 7%, X²(3) = 19.88, p = 0.0002].

In this study we found basal hypocortisolemia and blunted awakening salivary cortisol response in patients with treatment-resistant depression, compared to healthy controls. During dosing, 1 h 40 min after ingestion, we observed significant acute increases in salivary cortisol in the ayahuasca groups (C and MD), compared to placebo groups (C and MD). Moreover, 48 h after the dosing session (D2) with ayahuasca, patients' awakening salivary cortisol response is similar to levels detected in healthy controls. This was not observed in patients that ingested placebo, which continued showing different AUC of awakening salivary cortisol response compared to the control group. No significant differences were found in D2 regarding plasma cortisol.

In our recent randomized placebo-controlled trial we observed a significant antidepressant effect after a single dosing session with ayahuasca, when compared to placebo. Depression severity diminished significantly 1 day after dosing, and between-group differences increased from D1 to D7 (43).

Our results partially corroborate our initial hypothesis. We expected different levels of cortisol between patients and controls at baseline and, in fact, cortisol levels (both plasma and salivary) were reduced in the group of patients when compared to healthy controls. We also estimated that the cortisol levels will be correlated with the severity and/or duration of disease. Unexpectedly, cortisol levels (plasma and salivary) observed in the group of patients were not correlated with severity nor disease duration. Nevertheless, at baseline, a positive correlation between plasma and AUC of awakening salivary cortisol levels was found for controls, but not for patients.

Cortisol is a steroid hormone that triggers stress response in an adaptive way: it increases cardiovascular and respiratory activity, mobilizes glucose to provide enough fuel to remove the stressor and limit acute inflammation processes (45). Not only the excess, but also the reduction of cortisol is believed to be harmful. Major depression has been traditionally associated with hypercortisolism, but an increasing number of studies have found evidence of hypocortisolism in depression (8, 46, 47). Chronic decreased cortisol levels induce non-specific symptoms such as general malaise, weakness, low blood pressure, muscle weakness, loss of appetite and weight, gastrointestinal complaints and immunological dysfunction (48), some of which are very common in major depression. Besides depression, low cortisol levels have been found in some other conditions, such as Addison's disease (49), adrenal insufficiency (50, 51), and post-traumatic stress disorders (52).

In the literature, more severe depression, marked by chronic or recurrent disease episodes, is frequently accompanied by hypocortisolemia (8, 46, 47). These patients often exhibit long-term exposition to stressors, which in the early phase of the disease may induce chronic upregulated activity of the HPA axis and hypercortisolemia, followed by reducing cortisol to very low pathological levels (53, 54). In addition, evidence suggests that prolonged use of some antidepressants may also lead to increased expression of cortisol receptors and increased sensitivity of negative feedback, thus decreasing cortisol levels to under homeostatic values (55). Our patients were all treatment-resistant and used an average of 3.86 ± 1.66 different types of antidepressants, which could in turns induced hypocortisolemia. Low cortisol levels in depression have also been associated with maladaptive coping style (56) and unfavorable socioeconomic status (13, 57). Patients in this study had a particular profile: they were all from low socioeconomic status background, low educated and living in a low-income household, in a stressful environment. Previous studies have reported similar results. They found low levels of cortisol at awakening in patients with depression from the same region, compared to healthy controls (58), and with patients from Canada (58). Due to significant levels of poverty and scarce government investments in this region, this population usually is submitted to a long exposure of adverse events in life, and early in life they cope with precarious physical and physiological health, as well as with cumulative social and economic disadvantageous conditions.

The etiology of hypocortisolism is explained by various theories. One of them shows that a greater and decompensated sensitivity in negative feedback of the HPA axis deregulates cortisol secretion (54). Also, hypocortisolism is associate to adrenal insufficiency, a failure to produce cortisol and a decrease or increase in Adrenocorticotrophic hormone (ACTH) concentrations that depend on the type of failure, whether primary or secondary, respectively (48). New evidence also points to the participation of paracrine and autocrine messengers in adrenal failure (59). While several theories try to explain the etiology of hypocortisolism, it seems that the best approach to elucidate this pathophysiological process involves the integration of all these theories.

The regulation of cortisol is an important physiological aspect on the way of achieving biological health, since cortisol is an integrative hormone with large potential of body modulation, particularly involved in the etiology of depression, engaging the immune, and monoaminergic systems (53, 60, 61). Moreover, optimal levels of cortisol are necessary to induce neurogenesis, possibly due to its modulatory properties over brain-derived neurotrophic factor (BDNF) (62, 63). This relation seems to be an important factor in the etiology of depression, considering that the effectiveness of traditional antidepressants seems to be mediated by neuronal plasticity and neurogenesis (64).

The reference values for the diagnosis of corticosteroid insufficiency vary. Some studies define hypocortisolemia when cortisol levels are below 15 mcg/dL, others define it under 10 mcg/dL (65–67). Studies that evaluated the utility of basal morning serum cortisol measurements in the diagnosis of adrenal insufficiency showed that a threshold of 10 mcg/dL leads to 77% of specificity, and 62% sensitivity, defined by a subnormal serum cortisol response to insulin-induced hypoglycemia (68, 69). Therefore, in this study we considered 10 mcg/dL as cutoff reference value for true hypocortisolemia. We observed that 61% of our patients showed relative hypocortisolemia (<15 mcg/dL), and 22% true hypocortisolemia (<10 mcg/dL). On the other hand, awakening salivary cortisol response has been less used than plasma cortisol to monitor adrenal insufficiency, as it has not been fully validated as a diagnostic test.

Plasma cortisol levels were positively correlated with saliva cortisol levels at baseline for controls, but not for patients, in line with previous studies that observed a correlation between total plasma and salivary cortisol in healthy subjects (70). The absence of correlation in the groups of patients could be due to a malfunction of the HPA axis (70) or by changes in concentration of CBG (Cortisol Binding Globulin), its transporter protein in the plasma (71).

Moreover, we also supposed that ayahuasca, but not placebo, could acutely increase cortisol levels, after the dosing session, in patients and controls. The large increase in cortisol levels by ayahuasca found 1 h 40 after intake, for health and depressive volunteers, corroborated our hypothesis. Previous studies in healthy subjects have also reported increased cortisol levels during the acute effects of ayahuasca (72–74), N,N-DMT (75), psilocybin (76, 77), and Lysergic Acid Diethylamide (LSD) (78). One should bear in mind that our patients presented, in general, hypocortisolemia, and blunted awakening salivary cortisol response. It is reasonable to consider that subjects who took ayahuasca, which immediately increased cortisol levels, probably were acutely benefited by the ingestion of the ayahuasca, thereby leading to a direction of achieving hormonal homeostasis.

The modulation of the HPA axis by antidepressants depends on the type of antidepressant used and treatment duration, acute or chronic. Noradrenaline or serotonin (5HT) reuptake-inhibiting antidepressants, such as reboxetine and citalopram, acutely stimulate cortisol secretion in healthy volunteers, probably due the elevation in 5HT levels (79, 80). On the other hand, some antidepressants, as mirtazapine, acutely inhibits cortisol release, probably due to its selective antagonism at 5-HT₂ receptors (79). It is interesting to notice that the long-term effects of antidepressants are frequently opposite to the acute ones. In the long way, reboxetine up-regulates cortisol receptors function, repairs the disturbed feedback control and normalizes HPA axis. Mirtazapine, within 1 week, markedly reduces HPA axis activity in depressed patients (79, 80). If the patient is resistant to treatments, and uses antidepressants by years, the long-term effects could be disturbed and followed by an unfavorable physiological response, and as cited above, the chronic use of some antidepressant could induce hypocortisolemia.

Here, the acute increases of cortisol levels by the ayahuasca can be due the rise in serotonin induced by the N,N-DMT, and β-carbolinic alkaloids (31), likewise, other studies support the modulation by serotonin on the secretion of Corticotrophin Releasing Hormone (CRH) and ACTH both at the hypothalamic and pituitary glands (81).

After 48 h of the dosing session (D2), no changes in cortisol levels were observed comparing to baseline levels, within each group, and between treatments, neither in AUC of awakening salivary cortisol response nor in the plasma. These results were different than the claim of our initial hypothesis: previously, we expected an increase in cortisol levels (plasma and salivary) in D2. Individual analysis of AUC and PC at D0 and D2 showed a large individual variability. Some studies also faced with this large individual variability at baseline levels and response of cortisol (82). However, after 48 h from the dosing session, the AUC of awakening salivary cortisol response of patients that ingested ayahuasca, and not placebo, is similar to control that ingested ayahuasca. This similarity of AUC between controls and patients that were treated with ayahuasca points to a beneficial physiological modulation by ayahuasca on awakening salivary cortisol response. This finding probably corroborates with the improvement of several physiological systems, emotional and cognitive aspects that are regulated by cortisol (83, 84). Despite cortisol levels showed large inter and intra-individual variability, and as the area under the curve of awakening salivary cortisol consisted of three saliva samples collected in the early morning hours, some studies have been considering the awakening salivary cortisol response as more reliable biomarker than plasma. Awakening salivary cortisol response was demonstrated to be sensible to treatment by antidepressants (47, 85–88), probably because this index is less modulated by circadian clock and daily stressors than plasma cortisol levels (89).

Some studies with animal models of depression, rodents and non-human primates, observed positive antidepressant effects with the use of ayahuasca or its specific components (90–92). Using the recently validated translational animal model of depression (93), young marmosets (Callithrix jacchus) that showed hypocortisolemia, were treated with nortriptyline during 7 days, a tricyclic antidepressant, or with a single dose of ayahuasca. It was observed that ayahuasca increased fecal cortisol levels until 48 h after it ingestion, and presented more notable antidepressant effects than nortriptyline, since it reverted depressive-like behaviors and regulated cortisol levels faster and longer, when compared with nortriptyline (93).

Although there is great evidence that normalization of HPA axis function is essential for the successful treatment of depressive patients (89, 94, 95) differently of our hypothesis, herein changes in cortisol levels induced by ayahuasca were not correlated with improvement in depressive symptoms observed from D1 to D7 after it intake. We speculate that this fact is probably due to the high heterogeneity of patients and relative small sample of volunteers per group in this study.

The present study supports the involvement of the HPA axis in the physiopathology of depression and the modulation of salivary cortisol by a single session with ayahuasca, both in depressive patients and healthy volunteers. The cortisol acts in regulation of distinct physiological, cognitive and emotional pathways, and previous studies suggested that regulation of its levels to standard values is considered as an important part of depression treatment. Thus, we argue that ayahuasca should be further investigated as treatment of depression. Taking these findings in account, this work contributes significantly to support the return of clinical studies with natural psychedelics applied to mental disorders.

NG-C, BL-S, DdA, JM-d-O, EA, and FP-F designed the experiments; RdA and AG measured hormonal data; FF, FP-F, HO, AG, and ES collected experimental data, carried out statistical analysis and prepared the figures; NG-C, BL-S, DdA, FP-F, AG, and ES: prepared the manuscript.