Ashwagandha: Cortisol, Stress and Hormonal Balance
Ashwagandha has more human trial data than almost any adaptogen, a consistent effect on cortisol and stress scores, and a specific liver caution worth knowing about.
The Short Answer
Ashwagandha (Withania somnifera) is the best-studied adaptogen in the human literature, with dozens of randomised trials reporting reductions in perceived stress scores and in morning cortisol at doses of 300 to 600 mg per day of a root extract standardised for withanolides. Trials also report improvements in sleep quality, modest increases in testosterone in men, and small strength and body composition changes alongside resistance training. Two things temper the picture: many trials are small, short and conducted by parties with a commercial interest, and a genuine if uncommon signal of liver injury has led several regulators to require warnings.
What an Adaptogen Claim Actually Means
The adaptogen category originated in Soviet-era research and describes a plant compound said to increase resistance to stress without acting as a stimulant or sedative. It is not a regulatory or pharmacological classification, and it is used loosely enough in marketing to be nearly meaningless.
Ashwagandha is worth separating from the category because it has actual trial data. The active constituents are withanolides, a family of steroidal lactones, with withaferin A the most studied. Root extracts and leaf extracts differ substantially in withanolide profile, and this distinction matters for both effect and safety.
Traditional use in Ayurvedic practice is as a rasayana, a rejuvenating preparation, typically as root powder over extended periods. Modern supplementation uses concentrated extracts at doses that deliver considerably more withanolide content than traditional preparations, which is relevant when traditional-use safety history is cited in support of modern products.
Mechanism as Currently Understood
Ashwagandha appears to act on the hypothalamic-pituitary-adrenal axis, the system that governs cortisol release. Animal work shows reduced cortisol and reduced adrenal hypertrophy under chronic stress conditions, along with normalised stress-induced changes in neurotransmitter signalling.
Withanolides also have GABAergic activity, which is consistent with the sedative and sleep-related effects reported in trials, and there is evidence of effects on inflammatory signalling and on thyroid hormone output. The thyroid observation is a double consideration: trials have reported increases in thyroid hormone in participants with low output, which may be desirable or undesirable depending on the individual, and represents a real interaction rather than an incidental finding.
The honest position is that the mechanism is only partly mapped. The consistency of the human stress and cortisol findings is what makes the compound interesting, more than the precision of the mechanistic account.
The Evidence by Outcome
Stress and cortisol
This is the most replicated finding. Multiple randomised trials at 300 to 600 mg per day of standardised root extract report reductions in perceived stress scale scores and in morning serum cortisol relative to placebo over four to twelve weeks. Meta-analyses support a consistent effect on self-reported stress and anxiety measures. Effect sizes on subjective scales are moderate, and cortisol reductions in the range of twenty to thirty percent have been reported, though baseline variability in cortisol is substantial.
Sleep
Trials report improvements in sleep onset latency, sleep efficiency and self-reported sleep quality, both in people with poor sleep and in generally healthy participants. Effects are modest and consistent in direction. Evening dosing is the convention for this objective.
Testosterone and male fertility
Trials in men report modest increases in total testosterone, typically in the range of ten to twenty percent from baseline, with larger reported effects in men with low starting values or under high stress. Trials in men with fertility concerns report improvements in semen parameters. These effects are real but small, and they should not be confused with pharmacological hormone therapy.
Strength and body composition
Trials combining ashwagandha with resistance training report small additional gains in strength measures and lean mass relative to training alone. The effect is plausible as a downstream consequence of reduced cortisol and improved recovery rather than as a direct anabolic action.
Cognitive measures
Small trials report improvements in memory and executive function measures, and improvements in reaction time. The literature here is thinner and more heterogeneous than for stress outcomes.
The trial quality caveat
A large fraction of ashwagandha trials are small, short, conducted at a small number of centres, and funded by extract manufacturers. This does not make them wrong, and the direction of effect is consistent across many studies. It does mean the effect sizes should be read as upper estimates rather than as reliable point values.
Extracts, Standardisation and What to Buy
Ashwagandha is one of the few botanicals where specific branded extracts carry the trial evidence, which changes what "ashwagandha" means on a label.
- Root-only water-based extracts standardised to around five percent withanolides carry much of the stress, sleep and testosterone trial evidence, typically at 300 to 600 mg per day.
- Root and leaf extracts standardised to higher withanolide percentages, often around ten percent, are used at lower doses of 125 to 250 mg per day and carry their own trial set. Leaf material has a different withanolide profile including higher withaferin A.
- Unstandardised root powder at traditional doses of several grams per day delivers a much lower withanolide load and is not the material most trials used.
Practical implications: match the dose to the extract, since 600 mg of a five percent root extract and 600 mg of a ten percent root and leaf extract are different interventions. Prefer products stating withanolide content and part of plant. And note that root-only preparations are the traditional material, while leaf inclusion is a modern efficiency, which is relevant to the safety discussion below.
Safety: The Liver Signal and Other Cautions
Ashwagandha is generally well tolerated in trials, with mild gastrointestinal effects and drowsiness the most common reports. Two safety matters deserve more prominence than they usually get.
Liver injury. A series of case reports describes cholestatic liver injury associated with ashwagandha supplementation, typically appearing two to twelve weeks after starting, presenting with jaundice, itching and marked elevations in liver enzymes, and generally resolving after stopping. The absolute frequency appears low relative to the volume of use, but the association has been considered credible enough that some regulators, including in Denmark and more recently in other jurisdictions, have restricted sale or required warnings. The practical takeaway: anyone with existing liver involvement should avoid it, anyone taking it long term has a reasonable case for periodic liver marker checks, and any jaundice, dark urine, persistent itching or right upper abdominal discomfort is a reason to stop and seek clinical assessment rather than to wait.
Thyroid and autoimmune considerations. Ashwagandha can increase thyroid hormone output, which is unhelpful in hyperthyroidism and can interact with thyroid replacement dosing. It also appears to stimulate certain immune activity, which is a reason for caution in autoimmune conditions and a reason to avoid it alongside immunosuppressive therapy.
Other contraindications. It is not recommended in pregnancy, where traditional use includes abortifacient claims. It has sedative activity that can add to that of benzodiazepines, alcohol and other central nervous system depressants. And it may lower blood glucose and blood pressure, relevant alongside medication acting in the same direction.
Cycling, Duration and When to Stop
Most ashwagandha trials ran eight to twelve weeks, and the longest are around six months. That is the window with evidence behind it. Continuous use for years is common in practice and unsupported by trial data in either direction.
Given the liver signal and the absence of long-term data, a structured approach is defensible: use it for a defined period with a defined objective, typically eight to twelve weeks, reassess against that objective, and take breaks rather than keeping it as a permanent stack member. This is one of the few compounds where periodic discontinuation has a real rationale rather than being folklore.
Reassessment signals are usable here, unlike for many compounds. Perceived stress, sleep onset and sleep quality are perceptible within two to four weeks. If nothing has changed by six weeks at an adequate dose of a standardised extract, the compound has had a fair test. If it has helped, the question becomes whether to continue, cycle or address the underlying stressor, and that is a genuine decision rather than a formality.
The AEONNN Perspective
Ashwagandha sits at an interesting point in AEONNN's evidence framework. Its Evidence layer support for stress and sleep outcomes is stronger than for most botanicals, which earns it Level A or B weighting depending on the outcome in question, while its Safety layer carries a specific and unusual item: a credible hepatic signal with regulatory attention behind it. That combination is exactly why the Safety layer functions as a veto rather than as a weighting factor.
It maps across Hormonal Optimization and Vitality, Sleep and Circadian Regulation, and Cognition and Neuroprotection, which is a wide footprint for a single botanical and a reason it appears in many profiles. The Regulatory layer is live and moving here, since several jurisdictions have changed their position recently, and a member in one market may face different availability and labelling than a member in another.
This is also a compound where AEONNN's temporal logic argues against permanence. The trial evidence describes eight to twelve week periods, and Insight Protocol frames it that way, with a defined review point rather than an indefinite entry. Contingency regards it as a candidate for stress and travel contexts specifically, which is closer to how the evidence was generated than year-round continuous use.
Pillar Matrix mapping
Hormonal Optimization and Vitality, Sleep and Circadian Regulation, Cognition and Neuroprotection
Database Matrix layers
- Evidence Layer (PubMed, Cochrane, ClinicalTrials.gov)
- Meta / Consensus Layer (JAMA, BMJ, specialty society positions)
- Quality / Formulation Layer (ConsumerLab, Labdoor)
- Safety Layer (DrugBank, FAERS)
- Regulatory Layer (EFSA, FDA, EMA)
Frequently Asked
How much ashwagandha should I take?
Most trials used 300 to 600 mg per day of a root extract standardised to around five percent withanolides, or 125 to 250 mg per day of higher-concentration root and leaf extracts. Match the dose to the specific extract rather than to the word on the label.
Does ashwagandha really lower cortisol?
Multiple randomised trials report reductions in morning serum cortisol, with figures around twenty to thirty percent in several studies, alongside reductions in perceived stress scores. Cortisol is variable, so individual results differ, but the direction is consistent across trials.
Is ashwagandha bad for the liver?
A credible series of case reports describes cholestatic liver injury appearing two to twelve weeks after starting and resolving on discontinuation. It appears uncommon relative to how widely the supplement is used, but several regulators have restricted sale or required warnings. Anyone with liver involvement should avoid it, and jaundice, itching or dark urine is a reason to stop and seek assessment.
When should ashwagandha be taken?
For sleep objectives, in the evening. For daytime stress, morning or split dosing. Trials have used both schedules successfully, and drowsiness is the practical determinant for many people.
Does ashwagandha increase testosterone?
Trials in men report modest increases in total testosterone, commonly ten to twenty percent from baseline, with larger effects in men with low starting values or high stress. This is a small effect and not comparable with hormone therapy.
Should ashwagandha be cycled?
There is a reasonable case for it. Trials ran eight to twelve weeks, long-term data are absent, and a hepatic safety signal exists. Using it for a defined period against a defined objective, then reassessing, is more defensible than indefinite continuous use.
Who should avoid ashwagandha?
Anyone pregnant, anyone with liver involvement, anyone with hyperthyroidism or on thyroid replacement without supervision, anyone with an autoimmune condition or on immunosuppressive therapy, and anyone taking sedatives, since effects can add.
Is KSM-66 better than other extracts?
Branded standardised extracts carry most of the trial evidence, which is a genuine reason to prefer a material with published trials over an unspecified extract. It reflects where the research was done rather than demonstrated superiority over other well-standardised root extracts.
Evidence and review
Any dosage ranges cited here reflect the ranges used in published human trials, not personal recommendations. Evidence in this field moves, so this article is reviewed quarterly and carries its last-updated date above. Nothing here is intended as medical advice, and supplementation should be discussed with a qualified clinician, particularly alongside prescribed medication or an existing condition.