Adaptogens in Your Cup: Ashwagandha, Tulsi, and Rhodiola Science

"Adaptogen" is a term from 1940s Soviet pharmacology, coined to describe substances thought to raise non-specific resistance to stress. It is not a recognised pharmacological class today, and the European Food Safety Authority has rejected the health claims submitted under it.

That does not mean the plants do nothing. Several have real, documented pharmacology. It means the umbrella term is doing more work than the evidence supports, and the gap between what has been studied and what is sold is wide.

This page describes what has been studied. It is not medical advice, and it has not been reviewed by a clinician. Several of the herbs below interact with common medication, and the safety section is the part worth reading most carefully.

The stress axis these are said to act on

Under stress, the hypothalamus releases CRH, prompting the pituitary to release ACTH, which triggers cortisol release from the adrenal cortex. The loop is self-regulating in the short term. Chronic activation is associated with fatigue, disrupted sleep and impaired glucose handling.

The adaptogen hypothesis is that certain plant compounds modulate this axis rather than stimulating or sedating directly. It is a plausible framing. The evidence supporting it varies enormously by plant.

Ashwagandha (Withania somnifera)

What is studied: Withanolide A and related steroidal lactones show activity at GABA-A receptors in laboratory models. Human trials of standardised root extract report reductions in self-rated stress scales and in serum cortisol.

The caveats, which are substantial: the trials are short, mostly small, largely industry-funded, and use 300–600 mg of concentrated standardised extract. A spoonful of powder stirred into a drink is not that dose, and a hot-water infusion extracts only the water-soluble fraction.

Safety — this is the important part. There are documented case reports of ashwagandha-associated liver injury, and several national medicines regulators have issued warnings. It is contraindicated in pregnancy, where it has traditional use as an abortifacient. It can raise thyroid hormone levels, which matters if you have thyroid disease or take levothyroxine. It may potentiate sedatives.

Tulsi / holy basil (Ocimum tenuiflorum)

What is studied: eugenol and related compounds show antioxidant activity in vitro. Small human studies have looked at stress and metabolic markers with mixed and generally low-quality results.

What is claimed and not supported: immune enhancement. This appears on nearly every tulsi product and does not have human evidence behind it.

Safety: animal studies show anticoagulant activity. Discuss it before surgery and alongside blood thinners. Tulsi is pleasant as a daily infusion; treat the wellness claims attached to it as marketing.

Rhodiola (Rhodiola rosea)

What is studied: salidroside and rosavin, investigated principally for fatigue and cognitive performance under stress. Hung, Perry and Ernst's systematic review found the trial base small and methodologically weak, with results that do not consistently replicate.

Safety: may interact with antidepressants and other serotonergic drugs. This is the interaction most likely to matter in practice, because the overlap between people taking SSRIs and people buying stress supplements is large.

What a tea actually delivers

This is the point most articles skip, and it cuts both ways.

A hot-water infusion extracts the water-soluble fraction of a herb and leaves the rest in the leaf. Compared with a standardised capsule you get a lower and less predictable dose of whatever is active — which means less of the effect the trials measured, and correspondingly less of the risk described above.

If you want the effect a trial demonstrated, tea is not the delivery method that trial used. If you want a pleasant daily drink with a long traditional use and modest exposure, tea is exactly the right form.

Gentler alternatives with better-characterised evidence

If the goal is a calming evening drink rather than a pharmacological intervention:

• Chamomile — apigenin binds benzodiazepine receptor sites in laboratory work; long traditional use; caution with ragweed allergy.
• Lemon balm — rosmarinic acid inhibits GABA-transaminase in vitro; small trials of concentrated extract report reduced self-rated anxiety.
• Valerian — the most studied sleep herb, with genuinely mixed results and a distinctive smell; interacts with sedatives.
• Rooibos — caffeine-free, low tannin, no meaningful interaction profile, and simply a good drink.

The honest summary

These are plants with some real pharmacology, a thin and often poor-quality human evidence base, an unregulated supplement market around them, and — in ashwagandha's case — documented safety signals that get very little airtime relative to the marketing.

Drinking a tulsi or rhodiola infusion because you enjoy it is entirely reasonable. Taking a high-dose standardised extract because a label promised stress resilience deserves a conversation with a pharmacist first, particularly if you take anything else.

The rest of the shelf

Ashwagandha, tulsi and rhodiola dominate the marketing, but a wellness aisle carries several more. Briefly, and with the same standard applied to each.

Panax ginseng. Ginsenosides, studied for fatigue and cognition with results that are mixed and mostly from small trials. Genuinely stimulating for some people, to the point of disturbing sleep. Interacts with warfarin — reduced INR has been reported — and may lower blood glucose, which matters alongside diabetes medication.

Eleuthero, sold as Siberian ginseng, is a different plant entirely with a thinner evidence base. The shared name is a marketing inheritance, not botany.

Cordyceps. Almost everything sold is CS-4, a cultivated mycelium fermentation rather than the wild fungus. Exercise-performance trials exist and are small, short and inconsistent.

Reishi. Beta-glucans and triterpenes, with immune-modulating activity demonstrated largely in vitro. There are case reports of liver injury associated with powdered reishi, which is worth knowing before a daily habit.

Maca. Studied mostly for libido and mood in small, low-quality trials. Raw maca is goitrogenic in quantity; traditionally it is cooked, and most commercial powder is gelatinised for this reason.

Lion's mane. Hericenones and erinacines stimulate nerve growth factor in laboratory models. One small Japanese trial in older adults with mild cognitive impairment reported improvement that faded after stopping. It has not been replicated at scale, and everything sold on the strength of it is running well ahead of that single result.

Reading a supplement label

If you do decide to buy an extract rather than drink an infusion, the label tells you more than the front-of-pack claim.

"Standardised to 5% withanolides" beats "root powder". Standardisation means the maker measured the active fraction. Bare powder means they did not.

Named extracts are what the trials used. KSM-66 and Sensoril are different ashwagandha preparations produced by different extraction methods, and the studies on one do not automatically transfer to the other. If a product cites a trial, check that it used the same extract.

A "proprietary blend" hides the dose. Ingredients are listed in order but quantities are not given, so a blend can carry a token amount of the expensive ingredient behind a large amount of rice flour. Treat it as a reason not to buy.

Look for third-party testing. USP, NSF or Informed Choice marks mean an independent laboratory verified content. In the United States supplements need no pre-market approval for efficacy or content, and analyses repeatedly find products that do not match their labels, along with occasional heavy-metal contamination.

Interactions worth checking before you commit

Not exhaustive, and an absence here is not clearance. The general rule is that anything with a real pharmacological effect has real interactions, and the herbs marketed hardest are the ones people take alongside the most medication.

Liquorice deserves the emphasis it rarely gets. It appears in a great many herbal blends as a sweetener rather than as a headline ingredient, so people consume it daily without ever having chosen it. The health and safety reference lists it alongside the others with the drug classes attached.

Placebo, and why that is not a dismissal

Every outcome in this field is self-reported: stress scores, mood, perceived energy, sleep quality. Those are precisely the measures most sensitive to expectation, which is why the trial base is so hard to interpret and why industry funding matters so much.

It is worth being clear about what follows from that, because two conclusions are usually drawn and only one is right.

The wrong conclusion is that the effect is fake. If a warm, aromatic drink taken deliberately at the end of a working day makes someone feel calmer, they are calmer. Ritual, warmth and a ten-minute pause are a real intervention, and the fact that the plant may be contributing little does not make the outcome imaginary.

The right conclusion is narrower: you are probably paying for the ritual, so buy accordingly. A pleasant tulsi or chamomile infusion delivers the ritual at low cost and low risk. A high-dose standardised extract costs more, carries the interaction profile above, and has to clear a much higher bar to be worth it — one that, for most of these plants, the published evidence does not yet clear.

Sources

• Tandon N, Yadav SS. Safety and clinical effectiveness of Withania somnifera. J Ethnopharmacol 2020. 10.1016/j.jep.2020.113705
• Björnsson HK et al. Ashwagandha-induced liver injury: a case series. Liver Int 2020. 10.1111/liv.14393
• Hung SK, Perry R, Ernst E. The effectiveness and efficacy of Rhodiola rosea: a systematic review. Phytomedicine 2011. 10.1016/j.phymed.2010.08.014