Adaptogens and their impact on the central nervous system - molecular mechanisms of stress protection

The brain pays a price for chronic stress that can be measured, although it is smaller than popular texts suggest. In the largest imaging meta-analysis to date, the hippocampus of individuals with depression was found to be 1.24% smaller than that of healthy individuals, and in recurrent depression, it was 1.44% smaller (Schmaal i wsp., Molecular Psychiatry 2016). Adaptogens fit into this puzzle as substances that are supposed to shift the stress response rather than block a single receptor. The problem is that some mechanisms described online come from cell cultures or mice and have never been tested in humans. Below, we separate one from the other, provide doses from specific trials, and show where evidence simply does not exist.

KEY INFORMATION
• The hippocampus in depression is smaller by 1.24%, not by several percent (Schmaal et al., 2016).
• Ashwagandha 240-600 mg per day reduced morning cortisol by 23-27.9% in two randomized trials.
• Hericenones from lion's mane do not stimulate NGF production, contrary to popular belief (Mori et al., 2008).
• Eight cases of liver damage after ashwagandha have been reported, including 3 deaths (Philips et al., 2023).

What does chronic stress do to brain structure?

Chronic stress damages the central nervous system through three described pathways: through excess glucocorticoids, through disruption of glutamatergic transmission, and through low-grade inflammation. Volume differences visible in MRI are real but small. The ENIGMA meta-analysis on 1728 patients and 7199 controls showed a hippocampal difference of about 1.2% (Schmaal i wsp., 2016). The effect was driven by patients with recurrent episodes, and there was none in the first episode.

Popoli and colleagues gathered evidence that glucocorticoids released during stress alter glutamatergic signaling in the prefrontal cortex and hippocampus (Nature Reviews Neuroscience, 2011). Beurel, Toups, and Nemeroff showed that the relationship between inflammation and depression works both ways (Neuron, 2020). Not every patient with depression has elevated inflammatory markers, so anti-inflammatory treatment is not a universal answer here.

Allostatic load, described by McEwen and Stellar in Archives of Internal Medicine, is not a single measurement but the sum of regulatory system wear after many mobilization episodes (1993). The body pays not for the stress itself, but for the fact that the response to it does not shut off in time. Therefore, it is measured by a set of cardiovascular and metabolic parameters, not by a single cortisol measurement.

It is also important to remember what these MRI numbers represent. They are average differences between large groups, not results that can be read in an individual. In the ENIGMA meta-analysis, the severity of symptoms at the time of enrollment in the study was not associated with the volume of any structure. An age of onset below 21 years was associated with a smaller hippocampus, by 1.85%. The statement 'stress eats your hippocampus,' which circulates in popular texts, describes a real phenomenon on the wrong scale.

Which mechanisms of adaptogens have been confirmed in humans, and which only in animals?

Only two mechanisms have solid confirmation in human studies: a reduction in morning cortisol and an improvement in measured scales of sleep and anxiety parameters. The rest, such as the increase of neurotrophic factors, inhibition of microglia, or action on the GABA-A receptor, comes from cell cultures and rodents. Panossian and colleagues described adaptogens as substances with a biphasic effect, which behave like mild stressors in small doses (Medicinal Research Reviews, 2021).

This table looks more modest than a typical marketing description, and that is exactly how it should be. The leap from astrocyte cells to the human hippocampus is often overlooked in advertising texts, yet it determines whether a substance truly changes something in your brain.

Panossian points out another feature of this group: the biphasic response. In small doses, an adaptogen behaves like a mild stressor and activates the same adaptive pathways that protect the cell from strong stress. In large doses, the same extract may act oppositely. This model leads to a practical conclusion: more does not mean better, and exceeding doses from studies lacks pharmacological justification.

The second reason why results from test tubes do not translate to humans is concentrations. Receptor binding studies are conducted at concentrations in the range of tens of micrograms per milliliter in direct contact with tissue. After swallowing a capsule, the substance must pass through the intestine, liver, and blood-brain barrier, and how much reaches the neuron has not been measured for most adaptogenic extracts. The European Medicines Agency dedicated a separate dokument refleksyjny, which describes how to evaluate such preparations. This is not an approved therapeutic indication.

Do adaptogens lower cortisol in humans?

Yes, ashwagandha lowers morning cortisol, and this is the best-documented effect of the entire group. In the study by Lopresti and colleagues, 60 individuals with elevated stress took 240 mg of Shoden extract or placebo for 60 days. Cortisol dropped by 23% in the active group, while in the placebo group it increased by 0.5% (Medicine, 2019). The HAM-A anxiety scale significantly improved compared to placebo, while the DASS-21 scale result was only at the threshold of significance.

The second study, by Chandrasekhar and colleagues, used a higher dose of 600 mg per day in 64 individuals over the same period. The result on the perceived stress scale decreased by 44% compared to 5.5% in the placebo group, and serum cortisol decreased by 27.9% compared to 7.9% (Indian Journal of Psychological Medicine, 2012). A meta-analysis of nine trials involving 558 patients confirmed the direction of changes: the average difference for cortisol was 2.58 units in favor of ashwagandha (Arumugam i wsp., Explore 2024).

It is worth noting one caveat here. A systematic review of 52 trials on plants and the hypothalamic-pituitary-adrenal axis showed that the study designs varied so much that the authors did not calculate a common effect (Lopresti i wsp., Nutritional Neuroscience 2022). The direction of cortisol change is therefore reproducible, but its magnitude depends on the preparation, the dose, and how stressed the studied group was at the start.

A separate issue is DHEA-S, an adrenal hormone usually presented as a counterbalance to cortisol. Here, two studies from the same team yielded divergent results. In a study on stress, DHEA-S decreased along with cortisol, which the authors interpreted as a general dampening of adrenal activity. In a study involving overweight men over forty, the same extract increased DHEA-S by 18% and testosterone by 14.7%, while cortisol remained unchanged (Lopresti i wsp., American Journal of Men’s Health 2019).

The studied population thus alters the hormonal outcome to the opposite. The popular saying about "improving the cortisol to DHEA ratio" is not supported by either of these two studies, as in one both hormones decreased, while in the other cortisol did not change. If you plan to conduct laboratory tests, it makes more sense to look at morning cortisol alone rather than at a composite ratio.

What do studies say about BDNF, the nerve growth factor and neurogenesis?

Here, popular descriptions often diverge from the original works. In verifying the literature for this text, we noticed that the most frequently repeated statement about lion's mane mushroom contradicts the study it cites. Mori and colleagues examined extracts from four edible mushrooms on the human astrocyte line 1321N1. The extract from lion's mane indeed increased NGF production via the JNK pathway, but hericenones C, D, and E did not (Biological and Pharmaceutical Bulletin, 2008).

The authors stated this directly in the conclusion: the compounds responsible for stimulating NGF synthesis are not hericenones. The statement "hericenones stimulate NGF" has circulated in Polish internet for years and has no basis in this work. In mice fed lion's mane powder for 7 days, the level of mRNA for NGF in the hippocampus increased, which is already an animal result, not a human one.

The same applies to the work of Brandalise and colleagues, usually described as hippocampal regeneration in an Alzheimer's disease model. The study involved wild-type mice, meaning healthy ones, and showed an increase in excitatory neurotransmission in the synapse of mossy fibers with the CA3 area and better recognition memory (Evidence-Based Complementary and Alternative Medicine, 2017). This is an interesting result because it concerns a healthy brain, but it is not evidence of reversing neurodegenerative changes. In humans, none of the cited trials measured BDNF or NGF in serum, so claims of increased levels of these proteins in humans remain hypothetical.

Even if someone were to perform such a measurement, its interpretation would be difficult. BDNF measured from peripheral blood largely originates from platelets, and its relationship with concentration in the brain remains disputed. Therefore, studies on Hericium in humans measured what can be reliably measured: cognitive test results, not signaling proteins. Mori and colleagues used the HDS-R scale, while Pingali and colleagues measured reaction times in psychomotor tasks.

It is worth distinguishing between two groups of compounds from Hericium. Hericenones come from the fruiting body, while erinacines come from the mycelium, and most supplements on the market are produced from either the fruiting body or grain substrate with mycelium. Since in a 2008 study hericenones did not stimulate NGF synthesis, the mere fact of declaring them on the label guarantees nothing. A manufacturer that cites the content of hericenones as a sales argument refers to a feature whose relationship with neuronal action has not been demonstrated.

How do adaptogens affect the GABA and glutamate systems?

The data comes from electrophysiology and receptor binding studies, not from humans. A methanol extract from ashwagandha root inhibited GABA and TBPS binding while simultaneously enhancing flunitrazepam binding, which is a ligand for the benzodiazepine site, by 20-91% depending on the concentration (Mehta i wsp., Indian Journal of Medical Research 1991). In mammalian spinal cord neurons, the extract increased chloride influx even in the absence of GABA, and bicuculline and picrotoxin negated this effect.

The authors summarized this as GABA-mimicking activity. This is an important distinction against the popular description of a "partial agonist": the work shows interaction with the GABA-A receptor complex involving the benzodiazepine site, studied in tissue preparations. No one has repeated this measurement in humans.

Cannabidiol acts on the same receptor, but differently. In a study on oocytes of the African clawed frog with human GABA-A receptors, it turned out to be a positive allosteric modulator of alpha1-6 beta gamma2 subtypes, with its action not engaging the classical benzodiazepine site (Bakas i wsp., Pharmacological Research 2017). The strongest enhancement concerned receptors containing the alpha2 subunit. Reishi, on the other hand, prolonged sleep in rats after three days of oral administration, as recorded in EEG and EMG (Cui i wsp., 2012). Again: rodents, not humans.

What does this mean in practice? If you take ashwagandha in the evening and simultaneously take a sleeping pill that acts on the same receptor, the risk of cumulative effects is real despite the lack of studies in humans. The direction of action of both substances is consistent, and that is enough to inform your healthcare provider.

It is worth distinguishing this from claims about glutamate. The statement that adaptogens "restore the balance of glutamate to GABA" in the human brain has no basis in any measurement. The concentrations of both neurotransmitters can now be assessed using magnetic resonance spectroscopy, but no one has published such a study with an adaptogen. All knowledge about the impact of stress on glutamatergic transmission comes from animal models and studies on glucocorticoids, not from trials on plant extracts.

Do adaptogens inhibit inflammation in the brain?

We do not know this in humans. The hypothesis is reasonable, as the relationship between inflammation and depression is well documented, but none of the clinical trials on adaptogens measured microglial activation or inflammatory markers in the central nervous system. A systematic review with meta-analysis regarding the NLRP3 inflammasome included only animal models of depression and confirmed the involvement of this complex in the development of depressive behaviors (McColgan i wsp., Psychiatry Research 2026).

The fact that a given pathway exists in mice does not mean that a mushroom extract will turn it off in you. Beurel and colleagues point out another problem: it is unknown what percentage of untreated patients with depression actually have elevated inflammatory markers (Neuron, 2020). An anti-inflammatory intervention would therefore make sense for some patients, but not for all.

The practical conclusion is simple. If a seller describes reishi or cordyceps as "inhibiting neuroinflammation", they are referring to results from cell cultures and rodents, even if they do not specify this. Beta-glucans and triterpenes have documented immunomodulatory effects in such systems, but the translation to the human brain has not been measured. You can find more about the concept of adaptogens and their limitations in our text adaptogens and the nervous system.

Additionally, there is the quality of the animal studies themselves. The authors of the mentioned review assessed the risk of systematic error using the SYRCLE tool, and only 16 out of 23 qualified studies were included in the meta-analysis, covering a total of 170 animals. The groups are therefore small, and depression models in rodents replicate individual behaviors, not the disease.

This does not mean that the hypothesis should be rejected. It means that it stands on a different level of evidence than the reduction of cortisol, which has been measured in humans in two independent placebo-controlled trials. Distinguishing these two levels is the most practical skill when reading supplement descriptions. If a description mentions a molecular pathway without a clinical scale name and without the number of subjects studied, we are almost always talking about preclinical results.

In which CNS disorders do adaptogens have data from humans?

The list is shorter than usually stated and includes five areas. Beyond these, we enter a territory without clinical trials, where authors of texts typically fill the gap with a molecular mechanism. The following summary shows how many people have actually been studied in each indication.

The last line deserves a comment, as the previous version of this article mentioned PTSD in veterans as a documented indication. We did not find any work that confirmed this, and the identifier linked to that claim led to an article about periodontal cells in Marfan syndrome. We have removed the claim.

Pay attention to the numbers in the middle column. Aside from anxiety and stress, where a thousand participants were gathered, each indication is based on a single study involving 30 to 60 people. In pharmacology, this is a pilot level, after which a confirmatory study is normally planned, not a recommendation. None of these trials have been repeated by an independent team in another country.

The lack of trials does not mean that the substance does not work. It means that no one has checked it, and that is a completely different situation than a negative result. In the case of migraines, fibromyalgia, or post-traumatic stress disorder, descriptions on the internet create the impression of evidence, assembling a molecular mechanism with user testimonials. Such a setup can look convincing and means nothing. The practical test is simple: ask for the name of the scale, the number of subjects, and the comparison group. If any of these three pieces of information is missing, we are talking about a hypothesis.

What exactly did the trials show regarding depression, anxiety, and insomnia?

The strongest signal concerns anxiety and stress, the weakest depression. A meta-analysis of 12 randomized trials with 1002 participants showed a reduction in anxiety of 1.55 standardized mean differences and stress of 1.75, with doses of 300-600 mg per day. The authors immediately noted that the heterogeneity of the results reached 94% and that the certainty of the evidence is low (Akhgarjand i wsp., Phytotherapy Research 2022).

In depression, we have one direct comparison with a drug. Mao and colleagues administered rhodiola at a starting dose of 340 mg per day, sertraline 50 mg per day, or placebo to 57 patients for 12 weeks, with the possibility of dose escalation during the study. The reduction of symptoms after sertraline was greater, but the difference compared to placebo did not reach statistical significance in this trial, and rhodiola clearly caused fewer side effects (Phytomedicine, 2015). This is a pilot study, not a basis for discontinuing medication.

Insomnia was studied using actigraphy. Langade and colleagues administered 300 mg of root extract to 60 patients twice daily for 10 weeks. Sleep efficiency increased from 75.6% to 83.5%, while in the placebo group it increased from 75.1% to 79.7%, and sleep onset latency significantly decreased (Cureus, 2019). Sleep quality on the PSQI scale also improved more than after placebo. We have gathered practical tips regarding the timing and form of administration in a text about ashwagandzie na sen i stres.

The heterogeneity mentioned by the authors of the meta-analysis is not a technical detail. A value of 94% means that the differences between the results of individual studies mostly do not arise from chance, but from the fact that different things were studied: different preparations, different doses, different populations, and different scales. Therefore, the common average from such trials describes a trend, not an effect you can expect for yourself.

The dose-response analysis yielded another result. The authors observed a benefit regarding stress in the range of 300-600 mg per day, which is exactly where the doses from individual trials fall. Above this range, the curve did not rise. This is an argument against the popular practice of increasing the dose when nothing happens after two weeks.

Do adaptogens help with burnout and cognitive disorders?

For burnout, the data is weak, while for mild cognitive disorders, it is moderate. The work of Edwards and colleagues, often cited as evidence of the synergy between Rhodiola and ashwagandha, actually did not include ashwagandha. It was an open study, without randomization and without a control group, in which 101 people with life stress symptoms received WS 1375 extract from Rhodiola at a dose of 200 mg twice daily for 4 weeks (Phytotherapy Research, 2012). Improvement occurred in all seven questionnaires, but with such a design, it is impossible to distinguish the drug from expectation.

In mild cognitive disorders, the best-documented is lion's mane mushroom. A Japanese study with a double-blind trial included 30 people aged 50-80 who took 3 g of dried powder daily or placebo for 16 weeks. Results on the HDS-R scale increased at weeks 8, 12, and 16, and four weeks after the end of administration, they significantly decreased (Mori i wsp., Phytotherapy Research 2009). Laboratory studies did not show any adverse effects.

Ashwagandha has one study on cognitive functions in healthy volunteers: the aqueous extract improved reaction times in psychomotor tests (Pingali et al., 2014). Thirty people in one trial and twenty in another is too few to speak of proven pro-cognitive effects. We write more broadly about the mushroom in the text what lion's mane mushroom is.

In Edwards' study, there is a detail that clearly shows why the lack of a control group changes so much. Improvement occurred after just three days of administration. It is difficult to point to a biological mechanism that would shift the stress axis after 72 hours, and easy to point to the expectation of a participant who signed up for a study on a herb for stress. Without a placebo, these two explanations cannot be separated.

The decline in results four weeks after discontinuing lion's mane, described by Mori, indicates something different than a lasting rebuilding of neurons. If the extract were rebuilding connections, the effect would persist after the administration ended. The observed course rather fits a supportive action that disappears along with the substance. For someone considering supplementation, this means we are talking about continuous intake, not a treatment with lasting results.

How do adaptogens interact with psychiatric medications?

Most of the described interactions are theoretical, derived from in vitro studies, not measured in patients. This is not a reason to dismiss them, but to avoid stating percentages that no one has calculated. The previous version of this text claimed that ashwagandha enhances benzodiazepine sedation in 30-50% of cases. Such a study does not exist.

Inhibition of monoamine oxidase by Rhodiola was measured on methanol and aqueous extracts at a concentration of 100 micrograms per milliliter, where it reached 82-93% (van Diermen i wsp., Journal of Ethnopharmacology 2009). Such a concentration in the human brain after an oral capsule is unlikely, but in the case of concurrent antidepressant treatment, the decision is made by the attending physician, not the supplement seller.

It is worth mentioning separately the mood stabilizers. Lithium, valproate, and lamotrigine have a narrow therapeutic window, and no one has studied the interactions of adaptogens with them. This means that neither the direction nor the magnitude of any potential impact can be determined. In bipolar affective disorder, there is a second risk: substances with a stimulating effect can destabilize mood, while rhodiola is designed to increase drive.

The practical rule is as follows. Add the supplement to the same list where you keep your medications and show it at every visit. Doctors often do not ask about herbs because patients do not report them, and in the case of suspected adverse effects, the simplest explanation is then missing. Write down the name of the product, the daily dose, and the start date, because without these three pieces of information, assessing the causal relationship is impossible.

When should adaptogens not be used?

Four situations exclude the supplementation of adaptogens, the most serious of which concerns the liver. A team from several centers in India described 23 cases of liver damage after ashwagandha, including 8 after single-component preparations. The predominant picture was of cholestatic inflammation. Five patients had a prior chronic liver disease, three developed acute liver failure on this basis, and all three died (Philips i wsp., Hepatology Communications 2023).

Chemical analysis of the recovered preparations showed only natural phyto-compounds, without adulteration and without contamination. This is important because the popular explanation for such events usually sounds like 'it's the fault of a counterfeit product.' That was not the case in this series.

The remaining contraindications are as follows. Pregnancy and breastfeeding: women in this condition were excluded from all cited trials, so there is simply no safety data. Hyperthyroidism: in an eight-week trial involving 50 individuals with subclinical hypothyroidism, a dose of 600 mg per day raised T3 and T4 levels and lowered TSH, which works in the wrong direction in hyperthyroidism (Sharma et al., 2018). Acute psychosis and mania: this condition requires psychiatric treatment, and the supplement may delay proper therapy.

Additionally, common sense applies in autoimmune diseases. Medicinal mushrooms modulate immunity in both directions, and no one has measured this in patients on immunosuppressive drugs.

It is important to recognize the symptoms of liver damage, as they usually appear after a few weeks of use, not immediately. Yellowing of the whites of the eyes or skin, dark urine, pale stool, itching without rash, persistent fatigue, and pain under the right rib cage are signals that the preparation should be discontinued immediately and reported to a doctor. In the described series, the median time to symptom onset was 41 days, with a range from 14 to 540 days. Jaundice occurred in seven out of eight patients, and itching in five.

If you have a diagnosed liver disease, cirrhosis, or fatty liver with elevated transaminases, do not use ashwagandha at all. In this subgroup, the course was the most severe. For healthy individuals, a reasonable minimum is to measure ALT and AST before starting and after two months, especially when simultaneously taking paracetamol, statins, or antiepileptic drugs.

What doses and how long were used in the studies?

The doses from clinical trials are specific and fall within narrow ranges. The table below collects only those that were actually administered to people in the aforementioned studies, along with the duration of the study. Commercially available preparations may be weaker because they differ in standardization.

None of these trials lasted longer than 16 weeks, so nothing is known about safety and efficacy after a year. Cycles with breaks are a phytotherapeutic practice, not a conclusion from research. If you want to check if supplementation works for you, use the same tools that were used to measure the effect in the trials: the perceived stress scale PSS-10 and the anxiety questionnaire GAD-7. Both are free, have Polish versions, and take a few minutes. Fill them out before starting and after 8 weeks, as changes were demonstrated in that time window. You can find a comparison of the compositions and standardization of individual preparations in the category supplements.

When choosing a preparation, pay attention to three things on the label. The first is the form of the raw material: extract or ground root. Studies on cortisol were conducted on standardized extracts, not on powder, so 600 mg of powder and 600 mg of extract are two different doses of active substances. The second is the declared content of marker compounds, such as withanolides, given in percentages, not just a verbal description. The third is a certificate of analysis from an independent laboratory, covering heavy metals and microbiological purity.

The most common problem with labels is not false declaration, but the lack of one. The manufacturer provides the weight of the capsule and the concentration ratio but omits the percentage of active compounds, making it impossible to compare two preparations. If that number is missing, you cannot determine whether you are replicating the dose from the study or taking a fraction of it.

Frequently Asked Questions

Can adaptogens replace antidepressants?

No. In the only direct comparison, rhodiola at a dose of 340 mg per day performed worse than sertraline in terms of reducing depression symptoms, although it caused fewer side effects (Mao et al., 2015). The study involved 57 individuals with mild to moderate depression. This is too few to speak of a drug substitute.

How long does it take to see an effect on sleep and anxiety?

In trials on ashwagandha, measurement points were taken after 60 days for cortisol and anxiety scales, and after 10 weeks for sleep parameters from actigraphy. None of them measured the effect after a few days, so no quick action was demonstrated here. Promises of improvement within a week are not supported by data.

Does ashwagandha damage the liver?

For some individuals, yes. A team from India described 8 cases of drug-induced liver damage after single-component preparations, mainly with a cholestatic picture, and three patients with prior liver disease died (Philips i wsp., 2023). Chemical analysis did not show any adulteration of the preparations.

Does lion's mane regenerate neurons in humans?

There is no evidence for this in humans. A Japanese randomized study involved 30 individuals with mild cognitive impairment and showed improvement on the HDS-R scale after 16 weeks, but did not measure brain structure (Mori i wsp., 2009). The results declined 4 weeks after discontinuation.

Can ashwagandha be combined with sleeping pills?

Only after consulting with the attending physician. There is no pharmacokinetic study in humans that has evaluated this combination. Caution arises from in vitro work, where an extract from the root intensified the binding of flunitrazepam at the benzodiazepine site of the GABA-A receptor (Mehta i wsp., 1991).

Are adaptogens safe during pregnancy?

There is no data to confirm this, so the answer is no. Pregnant and breastfeeding women were excluded from all the trials cited here. In addition to the lack of data, there are cases of liver damage after ashwagandha reported in 2023 in the general population.

How do adaptogens affect the thyroid?

Ashwagandha raises T3 and T4 levels and lowers TSH. In an eight-week randomized trial involving individuals with subclinical hypothyroidism, the effect was beneficial (Sharma et al., 2018). In cases of hyperthyroidism, this effect is undesirable, so it is advisable to check TSH levels before starting and after 8 weeks.

If you are looking for products with specified standardization and analysis certificates, check the category suplementy i adaptogeny and the section herbs in the u Bucha store.

The article is informational and educational in nature and does not replace consultation with a doctor. If you are pregnant, breastfeeding, taking medications, or have chronic conditions, consult the use of supplements or herbs with a specialist.

Author: Michał Waluk · Opublikowano: 2026-05-11 · Aktualizacja: 2026-08-11