Ashwagandha for Stress in 2026: 491 Adults, Eight Weeks and a UK Safety Review Still Unfinished

Ashwagandha is being sold to stressed-out Britons as a calming herb while the government’s food watchdog still can’t say what a safe dose is. That isn’t a scare line. It’s the Food Standards Agency’s own position: “no safe levels established or set limits” for ashwagandha in food supplements.

And yet it’s everywhere. Gummies, powders, capsules, sleep blends, the thing your colleague stirs into an oat latte at 3pm. If you’re looking at ashwagandha for stress this autumn, the useful question isn’t whether it works at all. It’s how much it works, for whom, and what you’re agreeing to while the regulators finish their homework.

I’ve read through the main reviews, the NIH’s fact sheet and what Which? and the FSA have said. My view: the benefit looks real but modest, the studies are weaker than the marketing, and the safety question is open enough that waiting a few months costs you very little.

Ashwagandha for Stress: What the Trials Actually Measured

The best-known evidence comes from small randomised trials, and the US National Institutes of Health’s fact sheet gives a fair summary. A 2021 systematic review it cites included seven randomised trials covering 491 adults. Every one was run in India. They lasted six to eight weeks, mostly using root or root-and-leaf extracts at anything from 240mg to 1,250mg a day, and the benefits looked stronger at 500 to 600mg.

Those trials found lower stress and anxiety scores than placebo, less sleeplessness and fatigue, and lower blood cortisol. On paper, that’s a decent list.

Look at the footnotes though. The NIH points out that preparations vary wildly (different plant parts, different extraction methods, different standardisation), that many studies were done within traditional medicine settings, and that several outcomes were self-reported. One 120-person trial in overweight adults found less fatigue but no drop in perceived stress. Not every study agrees.

There’s a bigger gap that rarely gets mentioned. All seven trials in that 2021 review were run in India, in people who may differ from a stressed office worker in Leeds in diet, baseline health and what “stress” even means to them. The NIH makes the same point, noting that effects outside traditional medicine settings are unclear. I couldn’t find a large, independent UK trial, and I’m guessing there isn’t one, because nobody I’ve read cites it.

A newer pooled analysis, presented at a 2025 conference and listed in BJPsych Open, covered 15 studies and 873 people and reported lower cortisol, perceived stress and anxiety scores at eight weeks. I’d treat that one gently. It’s an abstract that the Royal College of Psychiatrists flags as not having gone through the journal’s normal peer review, so it’s a signpost, not a verdict.

Cortisol Is a Number, Not a Feeling

A lot of ashwagandha marketing leans on cortisol, the so-called stress hormone. The trials do report it falling. But a lower cortisol reading on a blood test isn’t the same as feeling calmer at your desk, and nobody should confuse the two.

The stress scores matter more, because they’re closer to what you care about. Those are questionnaires, though, which brings back the usual problem: people who think they’re taking something calming tend to report feeling calmer. Placebo groups tend to improve in stress trials, sometimes a lot.

So here’s a fair reading. Across the better-run studies, ashwagandha seems to beat placebo by a modest margin over about two months. Modest means you might notice it, not that it’ll rewire your Monday. If you’re hoping for the kind of effect you’d get from a proper intervention for anxiety, this isn’t it.

Who Paid for the Studies?

Which? makes a point that rarely appears on the front of a pot. The studies are small, often fewer than 100 people, they ran for only eight to twelve weeks, and many were funded by extract manufacturers.

That doesn’t make the results wrong. Industry-funded trials can be done well. But it’s a pattern we see across supplements, and it tends to flatter the product. When we looked at immune-boosting supplements for the school-run season, the same thing showed up: tidy results from small, funded studies, then a gap where independent replication should be.

Also worth knowing: ashwagandha has no authorised health claims in the UK. A brand can’t legally say it reduces stress on the label, which is why you’ll see vaguer wording like “supports calm” or “adaptogen”. The vagueness is the regulatory fingerprint.

The UK Safety Review Nobody Has Finished

This is the part that changes my advice.

In July 2024 the FSA opened a call for evidence on ashwagandha supplements, citing literature linking the herb to effects on thyroid hormones, blood sugar and possible liver toxicity. Responses closed on 2 September 2024. The agency said it would ask its Committee on Toxicity for a formal risk assessment, to work out whether a safe level can be set at all. The FSA’s page, last updated in November 2024, reports no outcome.

Which? says that review is still going and that a formal statement is expected later in 2026. I haven’t seen that statement published, so I can’t tell you what it says. I’m taking Which?’s word on the timing.

Meanwhile, other countries have moved faster. The NIH fact sheet notes that Denmark banned ashwagandha in 2023, and that France’s food safety agency ANSES advised against its use in certain groups in 2024. The reasons centre on liver injury reports, thyroid effects and concerns about pregnancy.

To be fair to the herb, the liver evidence is messy. The NIH describes case reports and a case series of acute liver injury, including jaundice, with most people recovering after stopping. But some patients already had liver disease or were taking other drugs, some products were blends, and the contents weren’t always independently checked. The FSA itself says it can’t yet confirm a direct causal link. And the NIH says it’s tolerated well for around three months, with mild stomach upset, loose stools and drowsiness the usual complaints.

But the sentence I keep coming back to is the NIH’s: long-term safety isn’t known. People buy ashwagandha as a daily habit. The trials ran for two months.

Ashwagandha for Sleep: The Side Door Into the Bedroom

A lot of people don’t buy it for stress at all. They buy it because they’re awake at 3am, and the pot says “restful”.

The sleep evidence is thinner again. The NIH describes a 2021 meta-analysis of five trials and 372 adults, lasting six to twelve weeks at 250 to 600mg a day, which found a small but statistically significant benefit. Effects were larger at 600mg, with at least eight weeks of use, and in people who actually had insomnia. Two individual trials are worth a mention: one in 150 adults with sleep problems, with sleep tracked by wrist-worn actigraphy, and one in 80 adults, half of whom had insomnia, where the benefit showed up mainly in the insomnia group.

Read that again. If you sleep reasonably well and just want a bit more calm, you’re not in the group where the benefit appeared. And the dose that did best is the top of the range, taken for two months or longer, which is exactly where you’d want the long-term safety data we don’t have.

Multi-ingredient blends make this worse, not better. When Which? tested 21 products for our look at menopause supplements, it turned into a 78-ingredient guessing game. A sleep blend with a pinch of five herbs tells you nothing about whether any one of them, at the dose in the trials, is doing anything.

“Adaptogen” Is a Marketing Word

You’ll see ashwagandha described as an adaptogen, a herb that supposedly helps your body adapt to stress. It’s a lovely idea. As far as I know it isn’t a regulated category or a medical one, and I couldn’t find an official UK definition. It’s a shelf label.

That matters because it lets products borrow credibility. A pot labelled “adaptogen” sounds like it sits somewhere between food and medicine, when legally it’s a food supplement with no authorised health claim. The FSA treats it as exactly that, and has been asking for manufacturers’ safety and stability data because it doesn’t have enough of it.

I’d put it bluntly. If a product’s pitch relies on the word “adaptogen” more than on a named extract, a stated dose and a published trial, put it back.

Who Should Skip It Outright

Which? and the FSA are clear on a few groups. Pregnant women should avoid it, and the FSA advises the same for anyone with an existing liver condition while the review continues. Which? adds anyone with an autoimmune condition such as lupus, multiple sclerosis or rheumatoid arthritis, because the herb may stimulate the immune system.

Thyroid is the other red flag. Small trials found it can raise thyroid hormone levels, which matters if your thyroid is underactive or overactive, or if you take thyroid medication. The NIH also lists possible interactions with diabetes drugs, blood pressure tablets, immunosuppressants and sedatives. Which? says the NHS flags antidepressants, blood thinners and statins too.

The NIH adds that ashwagandha may raise testosterone, so it’s considered potentially unsafe for men with hormone-sensitive prostate cancer, and that some experts advise against it in pregnancy and breastfeeding.

If you take any regular medication, speak to a pharmacist or GP before you buy a pot. I’d say that’s not boilerplate. For this herb it’s the main point.

What I’d Try Before a Pot of Powder

Here’s my contrarian bit, and I’ll stand by it. For most stressed people, ashwagandha is overrated compared with the dull stuff. Regular exercise, daylight in the morning and a sensible bedtime have far stronger and longer-running evidence than a two-month supplement trial from a single country.

Morning light is the cheapest of the lot, and we covered the evidence on SAD lamps last week. If you’re not moving much, even steady aerobic exercise (our piece on zone 2 cardio covers what counts) does more for stress than any capsule I know of. And if the dark months are what’s dragging your mood down, the NHS autumn advice on vitamin D is cheap, clear and doesn’t come with an unfinished safety file.

If your stress is bad enough that you’re reaching for supplements, that’s also worth a conversation with your GP. Anxiety that lasts weeks is treatable, and there are options with far better evidence than any herb.

If You’re Taking It Anyway

Plenty of people will. If you do, keep it sensible.

Choose a root extract from a brand that lists the extract type and the amount of active compounds, not a vague “proprietary blend”. The doses with the best trial support sit at 300 to 600mg of standardised root extract a day, which is what Which? cites. Don’t go above that because a gummy is tasty. Stop and see a doctor if you notice yellowing of the skin or eyes, which is the jaundice described in the liver case reports, or if anything else feels off.

Tell your GP or pharmacist you’re taking it, especially before any blood tests. Thyroid and liver checks are the ones that matter here, and a result that looks odd is far easier to explain when somebody knows about the supplement. Keep the pot, too. If a brand can’t tell you the extract type or the dose per capsule, the label is the first thing to check when something goes wrong.

Give it a defined run (say eight weeks) and then stop and see how you feel. Don’t let it become a permanent fixture by default.

The sensible position, for now, is that ashwagandha isn’t dangerous for most healthy adults over a short stretch, and it isn’t the stress cure the pots imply. It’s a mildly helpful herb with a question mark over it. When the Committee on Toxicity does report, that question mark will either shrink or turn into a rule. Would you rather find out which after you’ve bought a six-month supply, or before?