Ashwagandha: separating the evidence from the marketing

Ashwagandha evidence: what the research actually shows

Ashwagandha is one of the more credible adaptogens in the supplement world - not because the evidence is overwhelming, but because several well-controlled RCTs have shown meaningful reductions in perceived stress and serum cortisol. One trial found a 27.9% reduction in cortisol versus placebo over 60 days. That's worth taking seriously. It's also worth being honest about where the evidence runs thin.

What the evidence actually shows

The most-cited trial is probably Chandrasekhar et al. (2012), a double-blind RCT with 64 adults under chronic stress. Participants taking 300mg of KSM-66 ashwagandha root extract twice daily showed a 27.9% reduction in serum cortisol compared to 7.9% in the placebo group (p<0.0001). They also scored significantly lower on the Perceived Stress Scale (PSS) - a 44% reduction versus 5.5% in controls. That's a large effect size for a supplement trial. It's also a small sample, which matters.

A later meta-analysis by Pratte et al. - actually, more precisely, Salve et al. (2019) and a systematic review by Priyanka et al. (2021) both concluded that ashwagandha supplementation was associated with statistically significant improvements in stress, anxiety, and cortisol measures across multiple trials. But the review authors note the same thing I would: most studies are short (8-12 weeks), involve small samples (40-80 participants), and use varying extracts and doses. The signal is consistent. The confidence intervals are wide.

A 2019 RCT by Langade et al. (2019) with 60 participants found that 240mg of ashwagandha extract daily for 60 days reduced anxiety scores (Hamilton Anxiety Rating Scale) and morning cortisol significantly versus placebo. The cortisol reduction was statistically significant (p=0.0006). This is a lower dose than some trials, which is useful information if you're thinking about how much you actually need.

The biology: what ashwagandha is probably doing

Ashwagandha (Withania somnifera) contains a group of steroidal lactones called withanolides, along with alkaloids and saponins. The withanolides - particularly withaferin A and withanolide D - are thought to be the primary bioactive constituents, though the extract as a whole likely matters more than any single compound.

The main proposed mechanism involves the hypothalamic-pituitary-adrenal (HPA) axis. Under chronic stress, the HPA axis drives elevated cortisol output. Animal studies suggest withanolides may modulate this by influencing corticotropin-releasing hormone (CRH) signalling and reducing the sensitivity of the adrenal response. There's also evidence of GABAergic activity - withanolides appear to bind to GABA-A receptors, which may partly explain the anxiolytic effects seen in human trials.

Additionally, ashwagandha appears to reduce markers of inflammation - specifically C-reactive protein (CRP) and interleukin-6 (IL-6) - in some trials. Chronic stress and chronic inflammation are closely linked, so this isn't surprising. But I'd be cautious about overstating the anti-inflammatory angle; the human data on this is thinner than the cortisol data, and effect sizes in inflammation markers have been modest.

There's also emerging work on Nrf2 pathway activation - ashwagandha may support antioxidant defences by upregulating this pathway. That said, the human data on this is thin and I'd be overstating it to claim otherwise. Most of the mechanistic work is in cell lines or rodents.

Dosing: what the clinical evidence actually supports

This is where I see the most confusion in the supplement space. Doses in published RCTs range from 120mg to 600mg of standardised extract daily, with most positive results clustering around 300-600mg. Here's a rough breakdown:

• 240-300mg/day: Effective in the Langade et al. (2019) and Chandrasekhar et al. (2012) trials for stress and cortisol outcomes.
• 600mg/day: Used in the Choudhary et al. (2015) trial (n=52), which found significant improvements in memory, attention, and information-processing speed - though cognitive outcomes are a separate question from stress.
• Higher doses (up to 1250mg/day): Used in some sports performance trials, with modest effects on VO2 max and muscle recovery. The stress-specific evidence doesn't require these higher doses.

The extract standardisation matters enormously. KSM-66 (standardised to ?5% withanolides from root only) and Sensoril (standardised to ?10% withanolides from root and leaf) are the two most-studied proprietary extracts. Generic "ashwagandha powder" without standardisation is a different product in practice, even if the label looks similar. This is one reason why supplement labels lie - the difference between a standardised extract and raw powder can be enormous, and most labels don't make this clear.

I use 300mg of KSM-66 in the Kojo formula. That sits squarely within the range where the cortisol and stress data is most consistent. I could have gone to 600mg, but I didn't see evidence that doubling the dose meaningfully doubles the effect - and I'd rather not add cost or load without a clear reason.

Cortisol specifically: what the numbers look like

Cortisol is the most measurable outcome in ashwagandha trials, and it's where the evidence is clearest. Across trials, reductions in serum morning cortisol tend to fall in the range of 15-30% versus placebo in chronically stressed adults. That's meaningful. Morning cortisol is a reasonable proxy for HPA axis activity, and elevated chronic cortisol is associated with disrupted sleep, impaired immune function, and mood dysregulation.

A 2012 trial by Chandrasekhar et al. (2012) reported the 27.9% figure I mentioned above. A separate trial by Majeed et al. (2020) using a different extract formulation (Shoden, 240mg) found a 23% reduction in cortisol over 60 days in 60 adults, with significant improvements on the DASS-21 stress subscale (p=0.002).

What these trials don't tell us: whether the cortisol reduction is clinically meaningful for people with normal baseline cortisol. Most of these trials recruit people with elevated perceived stress scores at baseline. If your cortisol is already within a healthy range, the effect may be smaller or negligible. I think that's an honest caveat worth stating plainly.

Sleep: a secondary outcome worth noting

Several ashwagandha trials have included sleep quality as a secondary outcome, and the results are reasonably consistent. Langade et al. (2019) found significant improvements in sleep quality (Pittsburgh Sleep Quality Index) and mental alertness on rising. A dedicated sleep-focused RCT by Langade et al. (2020) - 60 participants, 120mg extract twice daily for 8 weeks - found significant improvements in sleep onset latency, total sleep time, and sleep efficiency versus placebo.

This makes mechanistic sense. If ashwagandha is genuinely modulating the HPA axis and has GABAergic activity, improved sleep is a plausible downstream effect. The GABAergic pathway in particular is relevant - GABA is the primary inhibitory neurotransmitter, and GABA-A agonism is the mechanism behind many pharmaceutical sleep aids (benzodiazepines, for instance). The effect size here is smaller and the doses studied for sleep tend to be lower, but the signal is there.

If sleep is your primary concern, it's worth reading my broader piece on supplements for stress and anxiety - ashwagandha is one piece of a larger picture.

Sports performance: real signal or overhyped?

There's a growing body of literature on ashwagandha and physical performance, and it's worth addressing because it often gets conflated with the stress evidence. The two are related but distinct.

A 2015 RCT by Choudhary et al. (2015) (n=50 healthy adults, 600mg KSM-66 for 8 weeks) found significant improvements in muscle strength (bench press and leg extension), muscle recovery, and testosterone levels versus placebo. The testosterone finding is interesting but modest - a mean increase of about 96 ng/dL from a baseline of roughly 630 ng/dL. Statistically significant (p=0.004), but whether that's physiologically meaningful in healthy young men is a fair question.

A meta-analysis by P�rez-G�mez et al. (2020) pooled data from five RCTs and found significant improvements in VO2 max and muscle recovery with ashwagandha supplementation. The authors note high heterogeneity between studies, which limits the conclusions. I think the sports performance data is interesting but not yet settled. The stress and cortisol data is more consistent.

Safety and tolerability: what the evidence says

Ashwagandha has a long history of use in Ayurvedic medicine, and the short-to-medium-term safety profile in clinical trials looks reasonable. Adverse events in most RCTs are mild and infrequent - primarily gastrointestinal (loose stools, nausea) at higher doses.

There are, however, a small number of case reports of hepatotoxicity - liver injury - associated with ashwagandha supplementation. Bj�rnsson et al. (2020) documented several cases of herb-induced liver injury linked to ashwagandha products. These are rare, but they're real. The mechanism isn't fully understood, and it's not clear whether the issue is the ashwagandha itself, contaminants in poorly manufactured products, or individual susceptibility. If you have pre-existing liver conditions, this is worth discussing with a doctor before supplementing.

Ashwagandha is also contraindicated in pregnancy based on animal studies showing potential abortifacient effects. The human data is absent - there simply aren't trials in pregnant women - so the precautionary position is sensible.

There are also theoretical interactions with thyroid medications (ashwagandha may influence thyroid hormone levels) and immunosuppressants. These are theoretical rather than well-documented, but worth noting if you're on either.

What ashwagandha probably won't do

I think it's worth being direct about the limits of the evidence, because the supplement industry rarely is.

Ashwagandha is not a treatment for clinical anxiety disorders or depression. The trials showing anxiolytic effects recruit non-clinical populations with elevated stress - not people with diagnosed GAD or MDD. The effect sizes, while meaningful in those populations, don't translate to "this replaces therapy or medication." If you're dealing with a clinical condition, please don't take supplement marketing as a substitute for proper care.

It's also not a reliable testosterone booster in the way some marketing implies. The testosterone findings exist, but they're modest, they're in specific populations (young men doing resistance training), and the clinical significance is unclear. The ashwagandha benefits that are best-supported are the stress and cortisol ones - the rest deserves more scepticism.

And it's not fast-acting. The trials showing the strongest effects run for 60 days. If you try it for two weeks and feel nothing, that's not necessarily a failure of the supplement - it may just be insufficient time. The HPA axis doesn't recalibrate overnight.

Frequently asked questions

My honest take

I've read a lot of supplement literature, and ashwagandha is one of the few adaptogens where I think the evidence is genuinely worth taking seriously. Not because the trials are large - they're not. Not because the mechanisms are fully understood - they aren't. But because the signal is consistent across multiple independent RCTs, the effect sizes on cortisol and perceived stress are meaningful, and the safety profile in the short term is reasonable.

What I'm less certain about: the long-term picture. We don't have good data beyond 3 months. The hepatotoxicity cases are rare but real, and I think anyone who tells you they're not worth mentioning is prioritising sales over honesty. I mention them because I'd want to know.

I'm also uncertain about who benefits most. The trials recruit chronically stressed adults with elevated baseline stress scores. If that's you, the evidence is more applicable. If you're functioning well and just looking for an edge, the data is less clear on whether you'll notice anything meaningful.

I included ashwagandha in the Kojo formula because I think the stress and cortisol evidence justifies it - not because it's fashionable. I chose KSM-66 at 300mg because that's the extract with the most clinical data at a dose that sits within the effective range. I didn't go higher because I don't think the evidence supports it for the specific outcomes I care about.

If you're sceptical of supplements in general, I think that's a healthy instinct. Most of the industry deserves it. But ashwagandha, used at a sensible dose of a standardised extract, is one of the cases where I think the evidence genuinely supports cautious optimism - not hype, just a reasonable expectation of modest, real benefit.