Adaptogens — What the Evidence Actually Shows

March 24, 2026 Tom 15 min read

Adaptogens evidence explained: what does the science say?

Some adaptogen research is genuinely compelling - ashwagandha reduced serum cortisol by 27.9% versus placebo in one 60-person RCT - but the category is also drowning in overclaiming. Here's what the primary literature actually supports, where the evidence thins out, and why the distinction matters if you're trying to make a sensible decision about what to take.

What the evidence actually shows

Let's start with the headline number everyone quotes. In a double-blind RCT by Chandrasekhar et al. (2012), 64 adults with a history of chronic stress were randomised to 300mg ashwagandha root extract twice daily or placebo for 60 days. The ashwagandha group showed a 27.9% reduction in serum cortisol (p < 0.0001) and significantly lower scores on the Perceived Stress Scale. That's a meaningful effect in a reasonably well-designed trial. I'm not dismissing it.

Rhodiola rosea has a reasonable body of evidence too. A systematic review by Hung et al. (2011) examined 11 RCTs and found consistent, if modest, effects on fatigue and stress-related symptoms - though the authors noted methodological heterogeneity across studies and called for larger, more standardised trials. The effect sizes were real but not dramatic. That's worth saying plainly.

Eleuthero (Siberian ginseng) has been studied since Soviet-era sports medicine, but a lot of that early research doesn't meet modern standards for blinding or reporting. More recent work is sparser. I'd be overstating it to call the human evidence for eleuthero robust in 2024. It's interesting. It's not settled.

What's biologically happening: the HPA axis and stress response

The term "adaptogen" was coined by Soviet pharmacologist Nikolai Lazarev in 1947. His definition was functional: a substance that increases non-specific resistance to stress. The modern mechanistic story is more precise - and more interesting.

Most well-studied adaptogens appear to act on the hypothalamic-pituitary-adrenal (HPA) axis, the system that governs your cortisol response to perceived threat. Under chronic stress, this axis becomes dysregulated - cortisol remains elevated beyond what's acutely useful, and downstream effects include disrupted sleep, impaired immune function, and changes in mood regulation.

Withanolides - the active steroidal lactones in ashwagandha - appear to modulate this response partly through inhibition of heat shock protein 90 (Hsp90), which is involved in glucocorticoid receptor signalling. Pratte et al. (2014) reviewed the pharmacology and noted activity at GABA-A receptors as a plausible additional pathway for the anxiolytic effects observed in human trials.

Rhodiola's active compounds - rosavins and salidroside - appear to influence serotonin and dopamine reuptake, and may modulate cortisol release via the sympatho-adrenal system. Panossian and Wikman (2009) proposed that adaptogens work partly by mimicking a mild stress signal, priming the body's stress-response systems without triggering a full cortisol spike. It's a plausible model. It's not the only one.

Worth noting: most of this mechanistic work is in vitro or animal models. The human data on exactly how these compounds affect HPA axis function is still being built out. The biology is coherent. The clinical translation is where it gets complicated.

Dosing: what the clinical evidence actually supports

This is where a lot of supplement marketing quietly falls apart. Doses used in successful trials are often higher than what ends up in products - particularly products using why supplement labels lie tactics to hide how little of each ingredient is actually present.

For ashwagandha, the most replicated effects in human RCTs come from doses of 300-600mg of a standardised root extract (typically KSM-66 or Sensoril), taken daily for at least 8 weeks. A meta-analysis by Priyanka et al. (2019) - covering five RCTs, 400 participants - found statistically significant reductions in stress and anxiety scores at these doses, with the effect becoming more pronounced beyond the 8-week mark.

For rhodiola, effective doses in trials have ranged from 200-680mg daily of a standardised extract (typically 3% rosavins, 1% salidroside). Below 200mg, the evidence is thin. Some products use 50mg and call themselves adaptogen formulas. That's not supported by the literature.

For context: at Kojo, I don't currently include ashwagandha or rhodiola in the formula - partly because the regulatory landscape for botanical health claims in the UK is genuinely complex post-Brexit, and I'd rather not make claims I can't back up under NHCR rules. What I do include are ingredients like Vitamin C at 500mg, which contributes to the reduction of tiredness and fatigue and to the protection of cells from oxidative stress - two functions that are directly relevant when the body is under sustained stress load. That's not the same as an adaptogen. But it's honest.

The cortisol question: can you actually measure a meaningful change?

Serum cortisol is a useful biomarker, but it's not as clean as it sounds. Cortisol follows a diurnal rhythm - it peaks within 30-45 minutes of waking (the cortisol awakening response, or CAR) and declines through the day. A single morning blood draw tells you something, but it doesn't capture the full picture of HPA dysregulation.

The better-designed adaptogen trials use multiple measurement points - morning and evening serum cortisol, or salivary cortisol across the day, or 24-hour urinary free cortisol. The Chandrasekhar trial I cited earlier used serum cortisol at a single time point. That's a limitation worth acknowledging, even though the effect was large.

A more rigorous approach came from Auddy et al. (2008), which measured serum cortisol at multiple time points in 98 subjects over 60 days using a standardised ashwagandha extract. They found dose-dependent reductions - the higher-dose group (500mg twice daily) showed greater cortisol reduction than the lower-dose group (125mg twice daily). The effect was statistically significant (p < 0.001) and consistent across measurement points. That's more persuasive to me than a single-timepoint study.

What I'd want to see more of: studies using the cortisol awakening response as a primary endpoint, and longer follow-up periods beyond 60 days. Chronic stress is a chronic problem. Eight-week trials are a starting point, not a conclusion.

Where the evidence is genuinely weak

I want to be direct about this, because the adaptogen category has a serious overclaiming problem.

Maca root is marketed heavily as an adaptogen. The human data on stress and cortisol is essentially absent. Most maca research relates to libido and menopausal symptoms, and even there the evidence is modest. Calling maca an adaptogen is a marketing decision, not a scientific one.

Schisandra chinensis has interesting preclinical data - it appears to modulate cortisol and adrenaline responses in animal models - but the human RCT evidence is sparse. Panossian and Wikman (2008) reviewed the available literature and found promising signals, but noted that most human studies were small, often unblinded, and conducted in the 1960s-80s under Soviet research conditions that don't meet modern standards. I find it interesting. I wouldn't stake a claim on it.

Ginseng (Panax) sits somewhere in between. There's decent evidence for cognitive fatigue reduction in some populations - a 2010 trial by Reay et al. (2010) found significant improvements in mental fatigue and working memory with 200mg Panax ginseng - but the cortisol and HPA axis data is less consistent. It's a useful herb with a real evidence base. Whether it fits the strict adaptogen definition mechanistically is still debated.

Adaptogens and sleep: an underappreciated connection

One thing that doesn't get enough attention: the relationship between elevated evening cortisol and poor sleep quality, and how some adaptogens may help by addressing the cortisol side of that equation rather than acting directly as sedatives.

In the Chandrasekhar trial, the ashwagandha group reported significantly improved sleep quality alongside the cortisol reduction. This makes mechanistic sense - if you're lowering the cortisol that's keeping the sympathetic nervous system activated in the evening, sleep architecture should improve. It's not a sedative effect. It's a downstream consequence of better stress regulation.

This is also where amino acids like glycine become relevant - though I want to be careful here. Glycine is included in the Kojo formula at 2,000mg, and there is interesting research on glycine and sleep quality, including a small RCT showing improvements in subjective sleep quality and daytime alertness. But large-scale human trials on glycine specifically for stress and cortisol are limited, and I wouldn't overstate the connection. The research is ongoing.

The quality problem: why most adaptogen products don't match the research

Even when the underlying evidence for an adaptogen is solid, most products on the market don't deliver what the trials used. There are three main failure modes.

Wrong extract. KSM-66 ashwagandha - a full-spectrum root extract standardised to 5% withanolides - is what most of the good trials used. Many products use a generic "ashwagandha root powder" with no standardisation. You have no idea what you're getting.

Wrong dose. As I mentioned above, effective doses in RCTs are typically 300-600mg for ashwagandha, 200-680mg for rhodiola. Products routinely use a fraction of this, especially in multi-ingredient formulas where the label looks impressive but each ingredient is dosed too low to do anything. If you're looking at supplements for stress and anxiety, dose transparency should be your first filter.

Wrong duration. Adaptogens are not acute interventions. Taking ashwagandha for a week before a stressful event is not how the evidence works. The RCTs showing cortisol reduction ran for 8-12 weeks minimum. If you're not prepared to take something consistently for two to three months, you're unlikely to see the effects the trials found.

Chronic psychological stress increases oxidative stress markers - reactive oxygen species accumulate faster than the body's antioxidant systems can neutralise them. This is a real physiological consequence of sustained HPA activation, not a vague wellness concept.

Vitamin C contributes to the protection of cells from oxidative stress - that's an authorised claim under NHCR, and it's supported by solid mechanistic and clinical evidence. At 500mg, the Kojo formula sits at a meaningful dose for this purpose. It's not an adaptogen. But if you're thinking about stress biology holistically, antioxidant capacity is part of the picture.

Plant polyphenols - including those found in grape seed and pine bark extracts - are also being studied for their effects on oxidative stress markers. The Kojo formula includes both. I want to be honest: large-scale human RCTs on these specific extracts for stress-related oxidative damage are limited, and the research is ongoing. The mechanistic rationale is coherent. The clinical evidence in humans is still being established.

How to actually evaluate an adaptogen claim

Here's a practical framework I use when I'm reading a new study or evaluating a product claim.

Is there a human RCT? Animal and in vitro data is interesting but not sufficient for a consumer decision.
What was the sample size? Studies under 30 participants should be treated as preliminary.
What extract and dose was used? If the product doesn't match the trial conditions, the trial evidence doesn't apply.
What was the primary endpoint? Perceived stress scales and cortisol are different things. Both matter. Neither is a proxy for the other.
How long did the trial run? Less than 4 weeks is almost certainly too short for HPA axis effects.
Who funded it? Industry-funded trials aren't automatically invalid, but they warrant more scrutiny. Look for independent replications.

By this framework, ashwagandha and rhodiola pass reasonably well. Most of the rest of the adaptogen category has significant gaps. That's not a reason to dismiss the field. It's a reason to be specific.

Frequently asked questions

Do adaptogens actually lower cortisol, or is that just marketing?

For ashwagandha specifically, there is genuine RCT evidence of serum cortisol reduction - 27.9% in the Chandrasekhar et al. (2012) trial (n=64, p < 0.0001). For most other adaptogens, the cortisol data in humans is much thinner. The claim is legitimate for ashwagandha at the right dose. It's marketing for most others. Chandrasekhar et al. (2012)

How long does it take for adaptogens to work?

Based on the RCT evidence, meaningful effects on stress and cortisol typically emerge after 8 weeks of consistent use. Some studies show earlier changes in subjective stress scores, but the cortisol biomarker data is more consistent at the 60-day mark. Don't expect acute effects. Priyanka et al. (2019)

Is it safe to take ashwagandha long-term?

The safety data from trials up to 90 days is reassuring - no serious adverse events in the major RCTs. There are rare case reports of liver injury associated with ashwagandha, which warrant caution in people with liver conditions. Long-term safety data beyond 3 months is limited. I'd cycle it. Chandrasekhar et al. (2012)

Can you take adaptogens with other supplements or medications?

Some adaptogens - particularly ashwagandha - may have additive effects with thyroid medications, sedatives, or immunosuppressants due to their activity at hormonal and immune pathways. If you're on prescription medication, check with a pharmacist or GP before adding any botanical supplement. This isn't fear-mongering; it's pharmacology. Pratte et al. (2014)

Is there a difference between ashwagandha root and leaf extract?

Yes, and it matters. Most RCT evidence uses root extract - specifically KSM-66 (full-spectrum root) or Sensoril (root and leaf). Leaf extracts have higher withanolide concentrations but less clinical data. Root extract is the better-evidenced choice for stress and cortisol outcomes. Check what's in your product. Auddy et al. (2008)

Are there any adaptogens with strong evidence for cognitive performance under stress?

Rhodiola rosea has the most consistent human data for cognitive fatigue specifically - particularly in situations of sleep deprivation or sustained mental demand. A trial in night-shift physicians found significant reductions in fatigue-related cognitive errors with 170mg rhodiola extract over two weeks. Panax ginseng also has reasonable evidence for mental fatigue. Reay et al. (2010)

My honest take

I've spent a lot of time in this literature. My overall read: ashwagandha is the real deal, at the right dose and extract, taken consistently. The cortisol data is more robust than I expected when I first started looking at it seriously. Rhodiola is genuinely useful for fatigue, with decent mechanistic support. Beyond those two, I'd want to see more before committing to a strong position.

What frustrates me about the adaptogen category isn't the herbs themselves - it's the way the marketing machinery has latched onto the concept and diluted it. Products with 50mg of ashwagandha in a "stress support blend" are not delivering what the trials showed. They're trading on the credibility of research conducted at doses ten times higher. That's the part I find hard to stomach.

I also want to be honest about where I've landed with the Kojo formula. I chose not to include ashwagandha or rhodiola in the current formula - partly regulatory caution, partly because I wanted to focus on ingredients where I could make specific, authorised claims backed by solid evidence. Vitamin C contributing to the reduction of tiredness and fatigue is a modest claim. It's also true, well-evidenced, and useful in the context of stress. I'd rather say something accurate and limited than something sweeping and vague.

The ashwagandha benefits literature is worth reading if you're considering adding it to your routine. Go in with realistic expectations, check the extract type, confirm the dose, and give it at least eight weeks. That's what the evidence supports. Everything else is noise.

References (9 studies)