Ashwagandha

What is ashwagandha?

Ashwagandha is the Indian-origin name of the plant Withania somnifera (L.) Dunal of the nightshade family (Solanaceae); in the Ministry of Agriculture and Forestry document, its Turkish name appears as “yalancı güvey feneri, aşvaganda.” The roots are widely used in Ayurvedic medicine; traditionally they are referred to as a tonic, hypnotic (sleep-inducing), sedative and diuretic, and an “adaptogenic” effect is reported as a known property. ^[7]

“Adaptogen” is a term from traditional use. What is measured in controlled human studies, however, is concrete outcomes: stress and anxiety scales, cortisol, sleep scores, hormone levels. This guide evaluates each claim by what these measurements show, in whom and over how long. ^{[1] [2]}

What is the difference between root and leaf extract, and what does standardization mean?

The roots of the plant contain ergostane-type steroidal compounds (withanolides) specific to the genus Withania; the Ministry document lists withaferin A and withanolide D as the main known compounds. The same document, based on EFSA’s 2012 plant compendium, notes that attention should be paid to steroidal lactones (withanolides) in the leaf and to piperidine alkaloids in the root. ^[7]

The Ministry document also states that withanone is 19 mg/g in leaves and 3 mg/g in roots on a dry-weight basis, and mentions as a possibility that undeclared leaf may have been added to root preparations in commercial products to increase withanone content. ^[7] A “root extract” and a “leaf or root–leaf mixture extract” are therefore not the same product.

Supplement extracts can be defined by withanolide content; this is the criterion behind the phrase “standardized root extract.” A hormone meta-analysis pooling 23 RCTs reports possible dose-response effects according to withanolide content; this suggests that the composition of the extract may affect the result, but it does not compare products. ^[4] A result measured in a single study applies to the extract in that study. According to the same review, Polish authorities limited daily intake in 2020, based on the DTU report, to 3 g of powdered root and at most 10 mg of withanolides; the review’s country table states that use in supplements is not regulated in Poland and is possible by notification. ^[6]

Does ashwagandha help with stress and anxiety?

The meta-analysis by Akhgarjand and colleagues took only RCTs: 12 articles, 1,002 participants in total, age range 25–48. Compared with placebo, ashwagandha significantly reduced anxiety (SMD −1.55; 95% CI −2.37 to −0.74) and stress level (SMD −1.75; 95% CI −2.29 to −1.22). Heterogeneity between trials was very high (I² 93.8% for anxiety; 83.1% for stress) and the authors rated the certainty of the evidence as low for both outcomes; they noted that higher-quality trials are needed. ^[1]

Another meta-analysis paints a more detailed picture: when seven trials for cortisol and six for the perceived stress scale (PSS) (488 participants in total) were examined, cortisol changed by −1.16 µg/dL (95% CI −1.64 to −0.69); but for perceived stress the difference was not significant (SMD −0.355; 95% CI −1.188 to 0.47). The authors state that this discrepancy calls for longer and larger trials. ^[2] That is, a drop in a biomarker does not always coincide with a drop in a person’s self-reported stress.

Both meta-analyses report outcomes such as scale scores and cortisol; they report no result for treating a diagnosed anxiety disorder. Study details for each of the benefits are covered on a separate page.

Detailed guideAshwagandha benefits: which claims are supported?→

Does ashwagandha improve sleep?

The systematic review by Cheah and colleagues took RCTs comparing ashwagandha extract with placebo in participants aged 18 and over; five RCTs (400 participants) were analyzed. A small but significant effect was found on overall sleep (SMD −0.59; 95% CI −0.75 to −0.42; I² 62%). The effect was more pronounced in adults diagnosed with insomnia, at doses of 600 mg/day and above and over 8 weeks or longer. Morning mental alertness and anxiety level also improved; there was no significant effect on quality of life. ^[3]

The same authors state that data on serious side effects are limited and that more safety data are needed to assess the safety of long-term use. ^[3] For other sleep-related options, there are melatonin and magnesium pages: melatonin and magnesium.

Cortisol, testosterone and thyroid: what is the effect on hormones?

The 2026 meta-analysis by Fornalik and colleagues covers 23 RCTs (1,706 patients) comparing oral ashwagandha with placebo in adults and reporting hormone levels. Cortisol fell (SMD −1.18). On the thyroid side, no effect was found for TSH and T3, and T4 rose slightly (mean difference 0.61 µg/dL). Testosterone rose in men (mean difference 57.43 ng/dL) and did not in women (5.09 ng/dL); the sex difference was significant. Estradiol did not change in either sex. The authors state that more standardized trials are needed because of heterogeneity between trials and limited data for some outcomes. ^[4]

These results are hormone levels; the clinical meaning of a certain rise in testosterone in a healthy man has not been shown in this meta-analysis. Data on clinical questions such as fertility and thyroid disease are much smaller and are covered on a separate page. There are also risk signals on the thyroid side: the Dutch RIVM has stated that current products may cause liver injury, thyrotoxicosis and adrenal suppression in susceptible people; which people are susceptible is not known. ^[6]

Is there evidence for muscle strength and exercise?

The study by Wankhede and colleagues is an 8-week, double-blind, placebo-controlled RCT: 57 young men aged 18–50 with little resistance training experience were randomized; the treatment group received 300 mg of root extract twice a day and the control group a starch placebo. The increase in bench press 1-RM was 46.0 kg in the treatment group and 26.4 kg in the placebo group; in leg extension, 14.5 kg and 9.8 kg. The testosterone increase was 96.2 ng/dL and 18.0 ng/dL. ^[8]

The limits of the study are clear: the sample is small, the participants are young, untrained men, and the duration is eight weeks. The authors also state that longer studies with women and older people are needed. The article reports in its acknowledgments that the root extract was supplied by two product manufacturers; the authors declare no conflict of interest. ^[8] A single RCT therefore keeps the evidence grade at D.

Can ashwagandha harm the liver?

The case series by Björnsson and colleagues reported five cases of liver injury attributed to supplements containing ashwagandha: three from Iceland (2017–2018) and two from the US DILIN network (2016). Three of the patients were men, and the mean age was 43 (21–62). In all, jaundice with nausea, fatigue, itching and abdominal discomfort developed after a delay of 2–12 weeks. The injury was cholestatic or mixed, itching and raised bilirubin lasted 5–20 weeks, and no hepatic failure developed. In four patients liver tests returned to normal in 1–5 months, and one patient was lost to follow-up. In the supplements the patients had used, ashwagandha was confirmed by chemical analysis and no other toxic substance was detected; four patients were also using additional supplements. ^[5]

The Ministry document reports that such cases have been reported in the literature since 2017 and summarizes 12 publications; it states that commercial herbal products can be mixed and mislabeled, so in each case it may not be clear whether the culprit is the plant or an ingredient or contaminant, but that in some cases the product was tested and no other contaminant was detected. The document states that the root and aerial parts of the plant are negative (N) on the Plant List and that maintaining this status was recommended. ^[7]

The summary table of the Ministry document also includes a series of 8 patients from India: in this series, which included people with prior liver disease, 3 patients who developed acute-on-chronic liver failure died during follow-up. ^[7]

The frequency is unknown; these series have no denominator. Case series are limited for establishing cause and effect, but the repeated pattern of jaundice, cholestasis and recovery after stopping the product has prompted regulators to act. Detailed case information, the series that ended in death, thyroid cases and drug interactions are covered on a separate page.

Detailed guideAshwagandha harms: liver, thyroid, pregnancy and interactions→

For whom is ashwagandha not suitable?

This list is not a clinical trial result but precautionary warnings from regulatory agencies. According to the review by Brendler and colleagues:

• France ANSES (April 2024): recommended that people with thyroid, liver or heart disease or hyperandrogenism, pregnant and breastfeeding women, people using sedatives and children up to 18 avoid these supplements. ^[6]
• Germany BfR (2024): asked in particular that children, pregnant women, breastfeeding women and people with liver disease avoid them, and that the general population also exercise moderation. ^[6]
• Netherlands RIVM (2024): recommended in particular that pregnant women not use it; one of its reasons is that, according to the WHO, the plant has historically been used to induce miscarriage. ^[6]
• Denmark DTU (2020): raised serious safety concerns because of the plant’s traditional use as an abortifacient, possible effects on sex and thyroid hormones, acetylcholinesterase inhibition, a possible effect on the immune system and five cases of liver injury, and concluded that a safe intake dose could not be determined from the available data. ^[6]

These warnings are precautionary. According to RIVM, there are no appropriate reproductive toxicity studies that would exclude the possibility of the plant harming the unborn child. ^[6]

Regulatory status: Türkiye and Europe

Türkiye. The Ministry of Agriculture and Forestry’s Commission on Plants That Can Be Used as Food re-evaluated the root of Withania somnifera; the opinion was adopted at the 22/12/2023 meeting. The root and aerial parts of the plant are negative (N) on the Plant List and it was recommended that this status be maintained; the reasons given were that no publication showing the plant’s use as food was found and that reports of liver toxicity exist. ^[7] On the Ministry’s Plant List, updated 16/07/2026, the root and aerial parts of Withania somnifera are also negative (N). ^[9] The review by Brendler and colleagues says that there is no entry for ashwagandha in Türkiye’s list of substances with restrictions on use in supplements. ^[6] Withania is also absent from the 21/09/2026 version of the Restricted Substances List. ^[10] Whether plants and herbal preparations can be used in foods is determined separately by the Plant List evaluated by the Commission on Plants That Can Be Used as Food; the status in this list is negative. ^[7]

Europe. In the EU, ashwagandha is regulated as a food supplement and national rules differ. The information below is reported from the review by Brendler and colleagues; among the authors of the review are people from an industry association, a food regulatory consultancy firm and a law firm. Denmark banned the use of ashwagandha as a food (supplement) in 2023, based on the DTU assessment. In the Netherlands, in 2025 an amendment was proposed that, if it enters into force, would ban its use in supplements. In June 2024, the Heads of Food Safety Agencies proposed ashwagandha to the European Commission for a priority procedure under Article 8 of Regulation 1925/2006; the review states that the procedure could lead to a ban if applied. ^[6]

This guide does not recommend a product or a place to source it.

How reliable is the evidence overall?

The studies underlying the stress, sleep and hormone results are small and short (for example 57 people, 8 weeks); the meta-analyses themselves report high heterogeneity and low certainty. ^[1] In the muscle strength study, the extract was supplied by manufacturers. ^[8] All five studies in the sleep meta-analysis were conducted in India and one was partly supported by a manufacturer. ^[3] Participants in the stress meta-analysis were aged 25–48; people under 18 and over 48 are not represented by these data. ^[1] On the safety side, DTU and BfR found the available data insufficient to determine a safe intake level. This guide does not recommend a personal dose.