Evidence Files

Ashwagandha: 3,000 Years of Ayurvedic Use and the First Decent Clinical Trials Are Actually Promising

Ashwagandha has been used in Ayurvedic medicine for 3,000 years. For most of that time, "used for centuries" was the only available evidence. In the last decade, a modest but real body of randomised controlled trials has accumulated. The results are more honest than the marketing — and more positive than most trending supplements deserve.

Published: 20 June 2026

Verdict

TREATMENT (partially confirmed)

We searched PubMed and Cochrane Library for ashwagandha evidence. Real randomised trials exist, which already separates it from most trending herbs — but their pooled results come with a caveat worth stating up front. A 2024 meta-analysis of five trials in 254 people found a significant reduction in anxiety scores (−5.96 points on the Hamilton scale) and improvements in several sleep measures, while reporting I² = 98%. That statistic means the trials disagreed with each other almost completely, and it is the single most important number in the ashwagandha literature. The authors' own conclusion is that results are promising but need larger studies to confirm.

Why it went from Ayurveda to European pharmacy shelves

Ashwagandha (Withania somnifera) is an adaptogen — a class of herbs claimed to help the body resist physical and psychological stress. The active compounds are withanolides, which appear to modulate the HPA axis (the stress response system involving cortisol).

The European supplement market discovered ashwagandha around 2015-2018 as interest in adaptogens grew alongside broader wellness culture. Unlike many trend-driven supplements, ashwagandha had a pre-existing research infrastructure from Indian academic medicine — which meant clinical trials were already being conducted.

What makes ashwagandha unusual in the supplement space is that some of its claimed effects have been tested in genuine RCTs with placebo controls, and the results aren't entirely disappointing — enough trials exist for two separate meta-analyses. This puts it in a different category from most herbs that ride wellness trends without any supporting data. It also means the limitations can be measured rather than guessed at.

What the clinical trials actually show

The trials are real, and that is genuinely unusual for a wellness-trend herb. What they add up to is a more interesting question.

Anxiety: a significant pooled result, with an enormous asterisk. The 2024 meta-analysis of five randomised trials in 254 participants found anxiety scores fell by 5.96 points on the Hamilton scale (95% CI −10.34 to −1.59) — and reported I² = 98%
What I² = 98% means, in plain terms. It measures how much of the variation between trials exceeds chance. At 98%, the studies are not really measuring the same thing: some found large effects, others little or none, and the average sits between them without describing any of them well. A pooled number from such heterogeneous trials is a starting point for research, not a basis for a recommendation
Sleep measures moved too. The same analysis found improvements in sleep onset latency, total sleep time and sleep quality scores — though not in wake-after-sleep-onset or time in bed. The same heterogeneity caveat applies
The authors' own verdict is cautious. They describe the results as promising while stating that larger studies are needed to confirm the effects on anxiety and insomnia. That is a fair summary, and notably more restrained than the packaging
A larger analysis tells the same story. A 2022 meta-analysis pooled 12 trials and 1,002 participants and found significant reductions in anxiety and stress — with heterogeneity of 93.8% and 83.1%, and the authors rating the certainty of the evidence as low for both outcomes. Two independent analyses, the same shape: significant on paper, inconsistent underneath
Sample sizes remain the limitation. Around a thousand people in total is small for a supplement sold to millions. By comparison, a single vitamin D trial settling its own questions enrolled 25,871
Extract standardisation matters. KSM-66 and Sensoril are the most studied preparations; raw powder products may carry different withanolide profiles, which makes transferring trial results to an arbitrary supermarket product unreliable

What the supplement industry claims vs. what trials show

Common marketing claims

"Reduces stress, boosts testosterone, builds muscle, sharpens focus, improves sleep, fights fatigue, balances hormones." Often positioned as a complete wellness solution.

What the RCTs actually support

Anxiety and sleep: a significant pooled result across five small trials, reported alongside I² = 98% — meaning the trials largely disagreed with each other. Testosterone, muscle and cognition: weaker still. The honest version is narrower than the label, and its authors say plainly that larger studies are needed to confirm it.

Frequently asked questions

Does ashwagandha actually reduce anxiety?

The pooled evidence says maybe, and says it quietly. A 2024 meta-analysis of five trials in 254 people found a significant reduction in Hamilton anxiety scores — while reporting I² = 98%, meaning the trials disagreed with one another almost entirely. The authors conclude the results are promising and need larger studies to confirm. That is a reasonable basis for trying it; it is not a basis for expecting it to work.

What does I² = 98% mean and why does it matter so much here?

I² measures how much the differences between trial results exceed what chance would produce. At 98%, the studies are effectively not measuring the same thing — some found large effects, others found little, and averaging them produces a number that describes none of them. It's the statistical equivalent of a group of people shouting different answers and reporting the mean. It doesn't mean the effect is absent; it means this body of evidence can't yet tell you its size.

Does it raise testosterone or build muscle?

These claims rest on considerably weaker ground than the anxiety and sleep ones — small trials, short durations, and a pattern where industry-funded studies report larger effects than independent ones. If the reason you're considering ashwagandha is muscle or testosterone, the evidence doesn't currently support that decision.

Is it safe?

Short-term use in trials has been generally well tolerated, but two caveats matter. Long-term safety data is thin, since most trials run 60–90 days. And there have been case reports of liver injury associated with ashwagandha products, which is a reason to stop and seek advice if you develop jaundice, dark urine or unexplained abdominal pain. It is also not a supplement to take during pregnancy.

If I want to try it, what should I look for?

Match the product to what was tested: standardised extracts such as KSM-66 or Sensoril are the ones that appear in trials, and raw root powder may carry a different withanolide profile entirely. Give it the duration the trials used — most ran 60 to 90 days — and be honest with yourself about what you're testing, since stress and sleep are exactly the domains where expectation does the most work.

Our Conclusion

Ashwagandha is one of the more honest stories in the supplement world, and the honest version is narrower than either the marketing or the sceptics allow. Genuine randomised trials exist, which most trending herbs cannot claim. The pooled result for anxiety and sleep is statistically significant. And it comes with I² = 98%, meaning those trials disagreed with each other almost completely — so the average describes none of them well. A second, larger analysis of 12 trials in 1,002 people found the same pattern and rated the certainty of the evidence as low. Around a thousand participants is also a thin basis for a product sold to millions; a single vitamin D trial enrolled over 25,000. The authors' own phrasing is the right one to carry away: promising, and needing larger studies to confirm. If you want to try it, use a standardised extract, give it the 60–90 days the trials used, and hold the result loosely — because in stress and sleep, expectation is doing at least some of the work no matter what's in the capsule.

Sources

Every claim in this article rests on a primary publication. Links go to the publisher's site or to PubMed, the US National Library of Medicine database.

This article is for informational purposes only and does not constitute medical advice. Trick or Treatment analyses the presence of clinical studies in open scientific databases — PubMed and Cochrane Library. The absence of studies in these databases does not automatically mean a drug is ineffective, but it does mean its effectiveness has not been confirmed by evidence-based medicine standards. Any treatment decisions should be made together with your doctor.