Ashwagandha

Ashwagandha is a root used in Ayurvedic medicine that has become the default "stress supplement" on every shelf — and for once the marketing points at something real. The strongest evidence is for stress and anxiety, where multiple controlled trials show lower stress scores and lower cortisol.¹⁵ Beyond that, the case thins out fast: sleep is promising but modest, the testosterone and performance claims rest on small studies, and its effect on thyroid hormone is a reason for caution, not a selling point. It also carries two things most product pages gloss over — a documented liver-injury signal and a firm contraindication in pregnancy. Here is the honest, benefit-by-benefit version.

What ashwagandha actually is — and what "adaptogen" does and doesn't mean

Ashwagandha is Withania somnifera, a small shrub whose root has been used in Ayurveda for centuries. The species name is a giveaway about its best-known effect: somnifera means "sleep-inducing." Modern supplements are made from a concentrated root extract (sometimes root-and-leaf), standardized to a group of steroidal compounds called withanolides — the constituents most researchers treat as the active fraction.

You will see it sold as an "adaptogen," a word worth being skeptical about. "Adaptogen" comes from mid-20th-century Soviet research and describes a substance that supposedly helps the body resist stress non-specifically. It is a traditional and marketing category, not a regulated pharmacological one, and slapping it on a label tells you nothing about dose, quality, or whether a given product does anything. So it is more useful to ignore the label and ask what has actually been measured in people. The answer: ashwagandha's clearest, most replicated effect is a reduction in self-reported stress and in the stress hormone cortisol.¹³ Everything else is a spectrum from "promising" to "barely studied," and the sections below grade each one.

The evidence, benefit by benefit

The "Evidence at a glance" grid above is the short version. Here is the reasoning behind each grade — including a caveat pattern you should carry into every ashwagandha claim you read: nearly every positive trial tested a specific branded commercial extract, most often KSM-66, Sensoril, or Shoden, at doses and preparations that differ from study to study.¹ That does not make the results fake, but it is why NIH concludes it is "difficult to identify specific extracts or recommended amounts."¹ Grade the body of evidence, not the single most flattering study.

Stress and anxiety

BModerate evidence

This is the reason to consider ashwagandha and the only claim that earns a B. Across several randomized, placebo-controlled trials and a 2021 systematic review, standardized extracts lowered scores on validated stress and anxiety scales compared with placebo.¹⁵ The effect is consistent enough, and mechanistically coherent enough (see cortisol, below), that the National Institutes of Health's Office of Dietary Supplements treats stress as its best-supported use.¹

The honest ceiling on that B: NIH describes these trials as "most of them fairly small in size and of short duration," and notes that because studies used different preparations and doses, it is hard to say which extract or amount to recommend.¹ Of the seven stress and anxiety trials NIH summarizes, six tested a branded commercial extract and participants were randomized for just 6 to 8 weeks.¹ None of that sinks the finding, but it is why this is a solid B and not an A. If a page tells you ashwagandha "melts away anxiety," it is overselling a real-but-modest effect.

Lowering cortisol

BModerate evidence

Cortisol is the measurable hook under the stress claim, so it gets its own grade — a slightly softer B. Multiple RCTs report lower serum cortisol on ashwagandha than on placebo.¹ The most-cited example is a 60-day trial in chronically stressed adults taking 600 mg/day of a full-spectrum root extract (KSM-66), which measured a 27.9% drop in serum cortisol from baseline, versus 7.9% in the placebo group.³ That is a real between-group difference, and it lines up with the subjective stress improvements. The same extract-variation and sample-size caveats apply, and cortisol is a noisy hormone that swings with time of day, sleep, and the blood draw itself — so treat the number as directionally meaningful, not a precise dose-response.

Sleep quality

CLimited evidence

Sleep sits right on the C/B border. A meta-analysis pooled by NIH's fact sheet found ashwagandha produced a small but statistically significant improvement in sleep, with the clearest benefit at 600 mg/day taken for at least 8 weeks.¹ "Small but significant" is the accurate phrase — this is not a sleeping pill, and the effect is easy to lose in the noise of an under-powered study. If your sleep problem is downstream of stress, the mechanism is plausible and the modest data supports a trial run. If you have a primary sleep disorder, this is not the tool.

Testosterone and male fertility

CLimited evidence

Here the internet gets ahead of the evidence. A few small trials in men do show movement: one 8-week RCT in aging men reported a 14.7% greater rise in testosterone and an 18% greater rise in DHEA-S versus placebo.⁴ NIH's review likewise notes ashwagandha "might increase testosterone levels."¹ But these are small studies of specific branded extracts, and the changes are modest and measured against a placebo group that also fluctuated. This is a genuine C: interesting signal, not an established effect you should expect in yourself, and not a substitute for anything clinical. The same NIH note flags the flip side — a possible testosterone bump is a reason for men with hormone-sensitive prostate cancer to avoid it.¹

Strength and exercise performance

CLimited evidence

A handful of small resistance-training and endurance studies report minor improvements in strength, muscle size, or VO2 max on ashwagandha. They are underpowered and — a recurring theme across this whole ingredient — heterogeneous in dose and extract, which is exactly the problem NIH flags when it says the varied preparations make specific recommendations difficult.¹ There is not enough here to call performance a reliable benefit, which is why it lands at C. If you take ashwagandha for stress and also lift, fine; do not buy it as an ergogenic aid.

Thyroid hormone

CLimited evidence

Read this one as a warning, not a feature. Ashwagandha can raise thyroid hormone levels (T3 and T4), and NIH lists thyroid-hormone medication among its interactions.¹ For someone with a sluggish thyroid that might sound like a perk, but it is unpredictable and unmonitored — and for anyone on levothyroxine or with hyperthyroidism, it is a way to tip toward too much thyroid hormone. The grade is C and the framing is caution: this is an interaction to manage around, covered in full under safety below.

KSM-66, Sensoril, or generic: what standardization actually buys you

Because the withanolides are treated as the active fraction, extracts are standardized to a withanolide percentage — the higher the number, the more concentrated the actives per milligram. This is the single most useful thing on an ashwagandha label, and it is why "300 mg" of one product is not the same as "300 mg" of another. Two branded extracts dominate the trials, and they are genuinely different:

• KSM-66 is a root-only extract standardized to more than 5% withanolides per 300 mg capsule.¹ It is the most-studied brand, used in most of the stress, cortisol, and testosterone trials cited above — including the 27.9% cortisol study.³ Its positioning leans toward daytime stress and performance.
• Sensoril is a root-and-leaf extract standardized to a higher withanolide content (commonly cited around 10%). Adding leaf raises the withanolide count but changes the constituent mix, and Sensoril is generally positioned for calm and sleep and dosed lower (often ~125–250 mg). Anecdotally it reads as more sedating, which is a feature at night and a nuisance by day.
• Generic or "whole root" powder — the cheap end — is often unstandardized, so you have little idea how many withanolides you are actually getting. Whole-root organic products (the traceability angle some brands sell) can be perfectly good, but "500 mg of root" tells you far less than "300 mg standardized to 5% withanolides."

What the percentage does not buy you is a guarantee of quality or safety. Withanolide content is a potency spec, not a purity certificate — it says nothing about heavy metals, correct species identification, or label accuracy, all of which have been problems in the herbal category. If you buy ashwagandha, the two things worth checking are the standardized withanolide percentage (so you can match a studied dose) and third-party testing (so the capsule contains what the label claims). We do not run affiliate links or product picks on this page; that is deliberate — this is a reference hub, and the buying guidance belongs in the reviews and comparisons linked below.

How much, and when

The dosage card above is the short answer; here is the context. Trials of standardized root extract for stress and anxiety have used a wide range — roughly 240 to 1,250 mg/day — with benefits appearing most consistently at 500 to 600 mg/day.¹ For sleep, the clearest results came from 600 mg/day sustained for 8 weeks or more.¹ A practical read: 300–600 mg/day of a withanolide-standardized root extract covers most of the useful evidence, and there is no good reason to megadose past it.

On timing, ashwagandha does not need to be tied to a clock the way caffeine does — the effect builds over weeks, not minutes. Because the drowsiness is real for some people (that somnifera name again), many take it in the evening, especially if sleep is the goal; if you feel foggy on a morning dose, move it later. Take it with food — the most common side effects are gastrointestinal, and a meal blunts them.¹ None of this is personalized advice; the ranges reflect what was studied, not what is right for your body or your medication list.

Safety: the parts most product pages soften

Ashwagandha is well tolerated by most healthy adults over the short term, and the everyday side effects are mild — loose stools, nausea, upset stomach, and drowsiness are the ones that show up in trials.¹ But "generally well tolerated" is where a lot of pages stop, and that is exactly where the important part begins. The interaction box and caution list at the top of this page are the structured version; this is the plain-spoken one.

Pregnancy: a contraindication, not a "check with your doctor"

Do not take ashwagandha if you are pregnant, trying to become pregnant, or breastfeeding. This is stronger than the usual supplement hedge because the concern is specific: some reports describe ashwagandha as having the potential to cause spontaneous abortion, and both NIH's professional fact sheet and the Merck Manual advise against use in pregnancy.¹⁷ The regulatory world has acted on it too — Denmark banned ashwagandha in food supplements in 2023, citing concerns that include effects on pregnancy and the reproductive system.⁸ There is genuine scientific debate about whether standardized root extract carries the same risk as traditional whole-plant preparations, and at least one herbal body argues the abortifacient reports have been overstated. In a situation where the downside is a lost pregnancy, the responsible call is the precautionary one: avoid it. For breastfeeding, there is simply no safety data — same answer.

The liver signal is real, and worth understanding before you start

Ashwagandha can, rarely, injure the liver. The NIH-run LiverTox database classifies it as a "likely cause" of clinically apparent liver injury (likelihood score B) — the first case was described in 2017, and reports have accumulated since.² Two things keep this in proportion. First, it is genuinely rare relative to how many people use ashwagandha; US usage has climbed sharply in recent years, into the millions, and injury remains uncommon.⁹¹⁰ Second, most documented cases followed a recognizable and recoverable pattern: injury typically appeared 2 to 12 weeks after starting, showed up as jaundice and itching (a cholestatic or mixed pattern), and resolved within 1 to 4 months of stopping.² The reassurance has a hard edge, though — rare fatal cases and cases requiring liver transplant have been reported, concentrated in more severe presentations.² Researchers using formal causality tools (RUCAM) have attributed specific cases to ashwagandha rather than coincidence.¹²

The practical takeaway is simple. If you develop yellowing of the skin or eyes, dark urine, pale stools, itching, right-upper-abdomen pain, or unusual fatigue while taking it, stop and seek medical care. Avoid ashwagandha altogether if you have liver disease or drink heavily, and be cautious stacking it with other supplements that can stress the liver. The full breakdown lives in our companion guide on ashwagandha side effects.

Interactions that actually matter

Ashwagandha is not a benign herb around medications. NIH and NCCIH flag several real interactions:¹⁶⁷

• Thyroid medication. Because ashwagandha can raise T3/T4, it can add to levothyroxine or push someone with an already-active thyroid toward too much hormone. If you take thyroid medication, this is a talk-to-your-prescriber situation.
• Sedatives and benzodiazepines. The sedative effect is additive — NCCIH notes supplements like ashwagandha may increase the effects and side effects of sedative drugs.⁶ Combining them (or ashwagandha and alcohol) can leave you more drowsy than you bargained for.
• Immunosuppressants. Ashwagandha can stimulate immune activity, which can work against immunosuppressant drugs — relevant for transplant recipients and people managing autoimmune disease.
• Antidiabetes and blood-pressure medication. It may add to the blood-sugar-lowering and blood-pressure-lowering effects of those drugs, risking levels that run too low.¹

Who else should skip it

Beyond pregnancy, breastfeeding, and liver disease, hold off or get clearance if you have a thyroid disorder, an autoimmune condition (Hashimoto's, lupus, rheumatoid arthritis — the immune-stimulating effect cuts the wrong way), or a hormone-sensitive prostate cancer.¹ Because of its sedative and central-nervous-system effects, the Department of Defense's supplement-safety program and standard surgical guidance advise stopping it about two weeks before surgery.¹¹ And there is not enough data to recommend it for children.

How long should you take it? The honest answer on cycling

Nobody knows the long-term safety, and pages that imply otherwise are guessing. The trials that establish ashwagandha's benefits and tolerability mostly ran 8 to 12 weeks; NIH summarizes it as well tolerated for up to about 3 months, with long-term safety unknown.¹ That is the edge of the map. Taking it for many months or years is not proven dangerous, but it is also not studied, and the liver signal is one reason not to assume open-ended use is free.

You will see a lot of advice to "cycle" ashwagandha — a few weeks on, a week or two off. That is a reasonable, conservative habit, but be clear that it is anecdotal, not evidence-based: there is no trial establishing an optimal cycle, and the practice is a sensible hedge against unknown long-term effects rather than a proven protocol. If you use ashwagandha, treating it as a time-limited tool — run it for a couple of months, reassess whether it is actually helping — is more defensible than taking it indefinitely on faith.

Who it's for, and who should skip it

Bottom line: ashwagandha is a reasonable, low-cost thing to try for stress and stress-linked poor sleep, where the evidence is real if modest and the short-term safety record in healthy adults is good. Buy a withanolide-standardized, third-party-tested extract, take 300–600 mg/day with food for six to eight weeks, and judge it on whether your stress actually improves — not on the hype.

Skip it entirely if you are or may become pregnant or are breastfeeding, if you have liver disease or drink heavily, or if you are on thyroid, sedative, or immunosuppressant medication without your prescriber's sign-off. And do not reach for it as a testosterone booster, a pre-workout, or a thyroid fix — the evidence for those is thin, and one of them is an interaction wearing a benefit's clothes. This is educational information, not medical advice; your clinician and pharmacist know your situation and your medication list, and this page does not.

Frequently asked questions