Benefits of Ashwagandha for Women: The Evidence, and the Liver Risk

The best-supported benefit is a measurable drop in cortisol, with modest improvements in sleep. But we have to be straight with you about the rest, because most articles on this are not: ashwagandha carries a LiverTox likelihood score of B — “likely cause of clinically apparent liver injury” — with roughly two dozen published cases, deaths in people who already had liver disease, and one woman who needed an emergency liver transplant. Denmark has banned it in supplements and the EU is reviewing it. It also raises thyroid hormones, which is a problem in a population where women are five to eight times more likely than men to have thyroid disease. This is a supplement with real effects — and real reasons for caution.

Key takeaways

• Cortisol reliably falls. Whether you feel less stressed is another question — a 2025 meta-analysis found the biomarker moved and perceived stress did not.
• Sleep effects are real but modest — the meta-analysis authors themselves call the effect “small but significant.”
• LiverTox rates it B: likely cause of clinically apparent liver injury. The damage is idiosyncratic, not dose-dependent.
• Denmark banned it. France advises against it. The EU is weighing a review.
• It raises thyroid hormones — helpful in subclinical hypothyroidism, dangerous in hyperthyroidism or on levothyroxine.
• The male fertility data does not transfer to women, and it is routinely repackaged as if it does.
• Almost every trial is 6 to 12 weeks. Most people take it indefinitely. Nobody knows what that does.

What it is, and why the extract on the label matters

Ashwagandha is Withania somnifera, a shrub used in Ayurvedic medicine and often called Indian ginseng, which it is not. The compounds usually credited are steroidal lactones called withanolides — though the NIH Office of Dietary Supplements notes that non-withanolide components may also be responsible for its effects, and LiverTox states plainly that the ingredient responsible for its activities and side effects is not known.

Root and leaf have different chemistry, and that distinction sits at the centre of the European regulatory dispute.

	KSM-66	Sensoril
Plant part	Root only	Root and leaf
Standardised to	≥5% total withanolides	≥10% withanolide glycosides
Typical trial dose	300 mg twice daily	125–250 mg daily

One correction worth making, because nearly every article gets it wrong: the claim that Sensoril is “twice as strong” comes from marketing, not from a head-to-head trial — we could not find one. And the two figures measure different things: total withanolides versus withanolide glycosides. Comparing 5% to 10% across those is not a like-for-like comparison.

Most positive trial results cluster at 300 to 600 mg a day.

What the evidence actually supports

Claim	Grade	The honest position
Lowering cortisol	Moderate	The strongest claim — but see the dissociation below
Sleep	Moderate-low	“Small but significant,” per the authors. Larger in diagnosed insomnia
Anxiety	Low-moderate	Two NIH bodies disagree with each other
Thyroid	Mixed — benefit and risk	Same mechanism, opposite outcomes depending on your baseline
Menopause symptoms	Weak	One trial, 100 women, 8 weeks
Female sexual function	Weak	One pilot study, 50 women
Female fertility	Insufficient	NIH: not enough evidence. The male data is separate
Exercise performance	Low	NIH: not enough evidence
Blood sugar	Low	A 3 mg/dL drop. Clinically trivial — and an interaction risk

Stress and cortisol — and the finding that complicates it

This is the best-supported claim. A meta-analysis of nine trials covering 558 people found a mean cortisol reduction of about 2.6 µg/dL, and a 2025 analysis found a significant drop of 1.16 µg/dL.

But that same 2025 analysis found no significant effect on perceived stress. A separate 120-person, 12-week trial found the same split — it reduced fatigue but did not reduce perceived stress.

So the honest summary inverts the usual claim: the biomarker moves, and how you feel may not. That is a genuinely important distinction for anyone buying this to feel calmer.

One more caution. One meta-analysis reported a standardised effect size of −1.75 for stress. An effect that large is implausible for any psychological intervention, and is better read as a signal of small-study or publication bias than as a selling point.

Sleep

A 2021 meta-analysis pooled five trials and about 400 participants. The overall effect on sleep was −0.59, which the authors themselves describe as “small but significant,” with better results in people with diagnosed insomnia, at 600 mg or more, and over at least eight weeks. It did not improve quality of life.

The US National Center for Complementary and Integrative Health puts it carefully: some preparations “may be effective” for insomnia and stress.

Where the two NIH bodies disagree

This is worth knowing because it tells you how thin the literature is. The Office of Dietary Supplements presents the anxiety evidence relatively favourably. NCCIH states flatly that “evidence is unclear about its effects on anxiety.” Two federal agencies reading the same studies reached different conclusions.

What the literature looks like structurally

Nearly all the positive human evidence shares a profile: 50 to 150 participants, six to twelve weeks, single-centre, conducted in India, using a branded extract supplied by the manufacturer, sometimes with manufacturer-affiliated authors. The 491 participants in one key systematic review were all from India. All five sleep trials were from India. All eight performance trials were from India.

ODS also notes that most studies were conducted within a traditional medical system, so effects when taken as a standalone supplement remain unclear. And effect sizes tend to shrink when trials run in Australia or the US.

None of that makes the research worthless. It does mean the confidence level should be “promising and under-tested,” not “proven.”

The thyroid question, which matters most for women

This is the part we most want you to read, because women are five to eight times more likely than men to have thyroid disease — and it is frequently undiagnosed.

The benefit side: a 2018 trial of 50 adults with subclinical hypothyroidism found 600 mg a day of root extract for eight weeks significantly improved TSH, T3 and T4 against placebo. One trial, 50 people, eight weeks — encouraging, not a basis for treating a thyroid condition.

The risk side is the same mechanism. Ashwagandha appears to push thyroid hormone up. There are three published case reports of thyrotoxicosis — dangerously overactive thyroid — in women aged 32 to 73, all resolving when they stopped. One was a 32-year-old woman whose TSH fell below 0.01 and who developed weight loss, tremor and palpitations six weeks in.

So whether raising thyroid hormone helps or harms depends entirely on where you started — which most people have not tested. NCCIH’s position is unambiguous: not recommended for people with thyroid disorders, and it may interact with thyroid medication. If you take levothyroxine, the effects are additive, and that is how you end up thyrotoxic by accident.

The liver question

LiverTox — the NIH’s reference database on drug-induced liver injury — assigns ashwagandha a likelihood score of B: “likely cause of clinically apparent liver injury.” That rating was updated in December 2024. Some articles still cite an older, milder score; B is current.

The pattern

• Onset two to twelve weeks after starting, though one 22-year-old woman presented after just a few days
• Usually cholestatic or mixed — jaundice and itching rather than a pure enzyme spike
• Resolution typically one to four months after stopping, with jaundice often protracted

What has been reported

Roughly two dozen published cases of clinically apparent liver injury. A 2020 series covered five cases in Iceland and the US, all of which resolved. A 2023 series from three Indian hospitals covered eight cases — and five of those patients had pre-existing chronic liver disease, three developed acute-on-chronic liver failure, and three died.

In 2023 a 41-year-old woman required emergency liver transplantation after two months of use.

The part that matters most

The doses involved were ordinary. Reported cases include 154 mg taken every other day. Others involved 450 mg once daily, and 2,100 mg a day for a year.

Which means this is idiosyncratic toxicity, not dose-dependent toxicity — it does not scale predictably with how much you take. So “I’ll just take a low dose to be safe” is not a valid safety strategy. There is no dose demonstrated to be free of this risk.

LiverTox’s operative instruction: use should be avoided in patients with cirrhosis or advanced chronic liver disease. Worth noting that fatty liver disease is common and frequently undiagnosed, so “I don’t have liver disease” is a less certain statement than it feels.

One caveat in the other direction, for fairness: many implicated products were multi-ingredient and potentially mislabelled. In several cases, though, the product was tested and contained ashwagandha with no contaminants.

Why Denmark banned it

Denmark banned ashwagandha in food supplements in 2023, following a risk assessment by its national food institute that cited traditional abortifacient use, effects on sex and thyroid hormones, immune effects and the liver cases — and concluded that no safe intake level could be established. Denmark then notified other member states.

Where things stand elsewhere: France’s food safety agency advised against use in pregnancy and breastfeeding and in people with endocrine, thyroid, liver or heart conditions. Sweden applies the Danish assessment case by case. Finland has considered a ban. The UK’s Food Standards Agency has a risk assessment underway. As of mid-2026, EU regulators are weighing a formal review that could produce restrictions, mandatory warnings or a ban.

The pushback, which is also worth hearing

Critics of the Danish assessment argue it failed to distinguish root extracts from leaf, leaned too heavily on animal data, and reviewed the literature selectively. That is a legitimate scientific objection rather than pure industry noise, and we would rather you knew both sides.

And a correction that cuts against the ban

The abortifacient concern that helped drive Europe’s most aggressive action traces largely to a monograph published in 2000. In 2024 the organisation that published it issued a statement saying its report had been misrepresented, and that there is no evidence ashwagandha root causes abortions.

That does not change the advice — do not take it in pregnancy, because there is no safety data in pregnancy at all, which is reason enough. But the specific evidentiary basis for one country’s ban is contested by the very source cited for it, and honest reporting means saying so.

Who should not take it

If you…	Why
Are pregnant or breastfeeding	No safety data exists. NCCIH says avoid
Have any liver condition	LiverTox: avoid in cirrhosis and advanced chronic liver disease. Deaths occurred in this group
Have a thyroid disorder or take levothyroxine	It raises thyroid hormones. NCCIH: not recommended
Have an autoimmune condition	Immunostimulant activity. Relevant to Hashimoto’s, lupus, RA and MS — all female-skewed
Have surgery scheduled	Sedative effects. NCCIH says stop beforehand
Take sedatives or benzodiazepines	Additive sedation
Take diabetes or blood pressure medication	Additive lowering — a hypoglycaemia risk, not a bonus
Take immunosuppressants or anticonvulsants	Documented interactions

Common side effects are mild and mostly digestive or sedative: stomach upset, loose stools, nausea and drowsiness.

One practical oddity worth knowing: ashwagandha interferes with digoxin and eleven other laboratory immunoassays, so it can distort blood test results. Tell your doctor you take it before any bloodwork.

The long-term gap

Both NIH bodies cap the evidence at “well tolerated for up to about three months,” and ODS states twice that evidence on safety over months or years is lacking.

Meanwhile the liver case series include a woman who took it for a year before developing jaundice. And most people who buy ashwagandha take it indefinitely.

That gap — between a three-month evidence window and open-ended real-world use — is the single most under-reported thing about this supplement.

What the regulation actually is

The FDA does not approve dietary supplements before they go on sale. Manufacturers self-certify safety and labelling, and the FDA can generally only act once a problem emerges.

Third-party seals help with a narrower problem — they verify identity, purity and manufacturing, not whether the product works:

• USP Verified — made in a certified facility, label accurate, contains what it claims. Does not test for banned substances.
• NSF Certified for Sport — every batch tested, including banned-substance screening.
• Informed Choice / Informed Sport — every batch tested for 230+ prohibited substances.

For ashwagandha specifically this matters, because LiverTox repeatedly notes implicated products were multi-ingredient and possibly mislabelled. A seal is your only practical defence against that failure. It does nothing about ashwagandha’s own effects on the liver.

Five things the other articles do not tell you

1. The cortisol drops. The stress may not.

A 2025 meta-analysis found a significant cortisol reduction alongside no significant effect on perceived stress, and a separate 120-person trial found the same dissociation. If you are buying this to feel calmer rather than to move a number on a lab report, that is the finding to weigh.

2. Liver injury does not track dose

Cases include 154 mg on alternate days and onset within days. It is idiosyncratic. There is no “safe low dose” strategy here, and the deaths occurred in people with pre-existing liver disease — a group that includes the large, largely undiagnosed fatty liver population.

3. The same thyroid mechanism is both the benefit and the harm — and the harm cases are all women

One trial used the T3 and T4 rise to help subclinical hypothyroidism. The published thyrotoxicosis case reports are in women aged 32 to 73. Whether it helps or hurts depends on a baseline most people have never measured.

4. The male data is being laundered into women’s content

The strongest body of ashwagandha evidence concerns male fertility and testosterone. It gets repackaged as “hormone balance” for women. But a 2025 sex-specific review found ashwagandha does not raise testosterone in women, and NCCIH says there is not enough evidence for female infertility, menopause or diabetes. The male data does not transfer.

5. Europe’s biggest regulatory action rests partly on a citation its own source disputes

The abortifacient concern behind Denmark’s ban traces to a 2000 monograph whose publisher stated in 2024 that it had been misrepresented. The pregnancy precaution stands on other grounds entirely — no safety data — but the specific evidence chain is contested.

So should you take it?

We are not going to give you a yes or a no, because it depends on facts about you that we do not have. What we would say to a friend:

• If you have a thyroid condition, liver condition, autoimmune condition, or are pregnant or breastfeeding — no. That is not a cautious hedge; those are the documented risk groups.
• If you take any regular medication, ask your doctor first. The interaction list is long and includes very common drugs.
• If you are healthy and want to try it for sleep or stress, use a third-party-certified single-ingredient product at 300–600 mg, give it eight weeks, and treat it as a trial rather than a permanent addition.
• Do not take it indefinitely on the assumption it is safe long-term. That evidence does not exist.
• Know the warning signs — yellowing skin or eyes, dark urine, itching, unusual fatigue, abdominal pain — and stop immediately if they appear.

And the unglamorous comparison worth making: for stress and sleep, the interventions with far stronger evidence and no hepatotoxicity are consistent sleep timing, exercise, and cognitive behavioural therapy for insomnia. Those are harder to buy in a bottle, which is precisely why the bottle sells.

Frequently asked questions

What are the benefits of ashwagandha for women?

The best-supported effect is a measurable reduction in cortisol, with modest improvements in sleep — the sleep meta-analysis describes its own finding as “small but significant.” Single small trials suggest possible benefits for menopausal symptoms and female sexual function, but each rests on one study of around 50 to 100 women over eight weeks. The NIH says there is not enough evidence for female fertility, menopause or blood sugar.

Is ashwagandha safe?

Not for everyone, and the risks are more serious than most articles convey. LiverTox rates it B — “likely cause of clinically apparent liver injury” — with roughly two dozen published cases, one emergency liver transplant, and deaths among people with pre-existing liver disease. Denmark has banned it in supplements and the EU is reviewing it. Avoid it entirely if you are pregnant or breastfeeding, or have liver, thyroid or autoimmune conditions.

Does ashwagandha actually lower stress?

It lowers cortisol — that is well replicated. Whether it makes you feel less stressed is less clear: a 2025 meta-analysis found a significant cortisol reduction alongside no significant effect on perceived stress, and a 120-person trial found the same split. Also note that NCCIH states evidence is unclear about its effects on anxiety, while the NIH Office of Dietary Supplements reads the same literature more favourably.

Can ashwagandha affect your thyroid?

Yes, and this matters particularly for women, who are five to eight times more likely to have thyroid disease. Ashwagandha raises thyroid hormones — helpful in one small trial of subclinical hypothyroidism, but there are three published case reports of thyrotoxicosis in women aged 32 to 73, all resolving on stopping. NCCIH does not recommend it for anyone with a thyroid disorder, and it interacts with thyroid medication.

Can ashwagandha damage your liver?

It can. LiverTox assigns it a likelihood score of B, meaning a likely cause of clinically apparent liver injury, with onset typically two to twelve weeks after starting and a usually cholestatic pattern. Critically, the injury appears idiosyncratic rather than dose-dependent — reported cases include doses as low as 154 mg taken on alternate days — so taking less is not a reliable safety strategy. Stop and see a doctor if you develop jaundice, dark urine or itching.

Why did Denmark ban ashwagandha?

Denmark banned it in food supplements in 2023 after a national risk assessment cited traditional abortifacient use, effects on sex and thyroid hormones, immune effects and liver injury cases, concluding no safe intake level could be established. France has since advised against it for several groups, and the UK has a risk assessment underway. Critics argue the Danish assessment did not distinguish root from leaf extracts and leaned heavily on animal data.

Can you take ashwagandha while pregnant?

No. NCCIH says it should be avoided during pregnancy and while breastfeeding. There is no safety data in pregnancy, which is sufficient reason on its own. Note that the widely repeated abortifacient claim traces to a 2000 monograph whose publisher stated in 2024 that its report had been misrepresented — but the precaution stands regardless, on the absence of evidence rather than the presence of it.

How long does ashwagandha take to work, and how long can you take it?

Trials generally run six to twelve weeks, with sleep benefits larger after at least eight weeks at 600 mg or more. As for duration, both NIH bodies cap the evidence at roughly three months, and the Office of Dietary Supplements states that evidence on safety over months or years is lacking. Since most people take it indefinitely, that gap between the evidence window and real-world use is the most under-reported thing about it.

What is the difference between KSM-66 and Sensoril?

KSM-66 uses root only and is standardised to at least 5% total withanolides, typically dosed at 300 mg twice daily. Sensoril uses root and leaf and is standardised to at least 10% withanolide glycosides at 125 to 250 mg daily. The common claim that Sensoril is twice as strong comes from marketing — we found no head-to-head trial, and the two percentages measure different compounds, so they are not directly comparable.

Does ashwagandha help with menopause symptoms?

One trial suggests it might. A 2021 study of 100 perimenopausal women taking 300 mg twice daily for eight weeks found significant reductions on the Menopause Rating Scale, along with increased estradiol and decreased FSH and LH. That is a single 8-week trial, and NCCIH’s position is that there is not enough evidence for menopause. The estradiol signal is also a reason for caution in anyone with hormone-sensitive cancer.

Does ashwagandha balance hormones in women?

That phrase mostly comes from male data being repackaged. The strongest ashwagandha evidence concerns male fertility and testosterone, and a 2025 sex-specific systematic review found it does not raise testosterone in women. NCCIH says there is not enough evidence for female infertility, and PCOS and PMS claims rest on narrative reviews rather than adequate trials. Treat “hormone balance” marketing with scepticism.

Related content

• Best perimenopause supplements
• Best magnesium supplement for women
• Creatine for women: what the evidence says
• All Culture Mom wellness writing