Substance and claimed effects

"Adaptogen" is a 1947 Soviet pharmacology coinage (Nikolai Lazarev) describing plant compounds that purport to non-specifically increase resistance to physical, chemical, or biological stressors without disrupting normal physiological function. The modern operational definition, formalised by Panossian and colleagues, is a substance that (i) modulates the hypothalamic–pituitary–adrenal (HPA) axis and sympatho-adrenal system, (ii) enhances stress-stimulus tolerance without producing dependence or tolerance, and (iii) shows multi-target rather than receptor-specific pharmacology Panossian 2017. The label is contested — most Western regulators do not recognise "adaptogen" as a pharmacological class — but the three botanicals covered in this entry (Withania somnifera/ashwagandha, Rhodiola rosea/golden root, and Hericium erinaceus/lion's mane) are the most-studied of the supplements marketed under the banner. They are typically taken as standardised extracts in capsule form, once daily, for at least 8 weeks. Claimed consequences across the catalogue's dimensions: lower cortisol and perceived stress, reduced anxiety and depressive symptoms, improved sleep onset and quality, more daily energy and resilience to mental fatigue, sharper cognition (memory, attention), and — in the case of ashwagandha specifically — modest male endocrine effects (testosterone, DHEA-S). The entry covers all three botanicals as a class, because at consumer-supplement level they are sold, stacked, and consumed interchangeably; per-botanical evidence and mechanism is documented in §3b.

Evidence by addressing question

Mechanism

Ashwagandha. The root contains ≥40 steroidal lactones (withanolides), of which withaferin A and withanolide D appear to drive most pharmacological activity Jamnekar et al. 2025. Withanolides cross the blood-brain barrier and act on at least three pathways: (1) HPA-axis dampening — preclinical models show reduced stress-induced corticosterone surges and restored glucocorticoid-receptor sensitivity, which translates clinically to 23–33% serum cortisol reductions in stressed adults at 600 mg/day Salve et al. 2019; (2) GABAergic action at GABA-A receptors, mechanistically analogous to (but far weaker than) benzodiazepines, with no dependence signal; (3) modulation of serotonergic and dopaminergic tone, plus NF-κB inhibition and Nrf2 activation contributing to anti-inflammatory and antioxidant effects Jamnekar et al. 2025.

Rhodiola. The active compounds are salidroside and rosavins (rhodiola is standardised to ≥3% rosavins and ≥1% salidroside in clinical-grade extracts such as SHR-5). Mechanism is dual: monoamine modulation (mild MAO-A and MAO-B inhibition, increased serotonergic/dopaminergic/noradrenergic signalling) and an Hsp72/NPY/HSF1-mediated cellular stress response pathway documented in Panossian's molecular work Panossian 2017. The single-dose stimulant-like effect maps onto the sympatho-adrenal system; the repeated-dose anti-fatigue effect maps onto HPA-axis normalisation with cortisol-awakening-response attenuation observed at 576 mg/day in stress-related fatigue Olsson et al. 2009.

Lion's mane. Two distinct compound families: hericenones (from the fruiting body) and erinacines (from the mycelium, with erinacine A as the lead compound). Both are low-molecular-weight diterpenoids that cross the blood-brain barrier and stimulate nerve growth factor (NGF) synthesis in astroglial cells; preclinical work also shows brain-derived neurotrophic factor (BDNF) induction and TrkA/Erk1/2 pathway activation promoting neurite outgrowth. The mechanism is genuinely distinct from ashwagandha/rhodiola — lion's mane is not a classical HPA-axis adaptogen but a neurotrophin promoter, and is grouped with adaptogens by marketing convention more than pharmacology.

Evidence

Ashwagandha — strongest of the three. A 2021 PLOS ONE meta-analysis of five RCTs (n=400) found a moderate-quality, statistically significant effect on overall sleep (SMD −0.59, 95% CI −0.75 to −0.42), with stronger effects in adults with diagnosed insomnia, at ≥600 mg/day, for ≥8 weeks Cheah et al. 2021. Subgroup effects: sleep-onset latency SMD −0.53, total sleep time SMD −0.45, sleep efficiency SMD −0.68. A 2025 BJPsych Open meta-analysis of 15 RCTs (n=873) confirmed an 8-week PSS reduction of −4.88 points and a cortisol reduction of −2.36 µg/dL Bachour et al. 2025. The Salve 2019 dose-finding trial established the canonical 600 mg/day dose, showing significantly larger PSS and serum cortisol reductions versus 250 mg/day Salve et al. 2019. Langade 2019 found significant improvements in sleep-onset latency, sleep efficiency, and HAM-A anxiety scores at 600 mg/day over 10 weeks in insomnia/anxiety patients Langade et al. 2019. Lopresti 2019 reported 18% testosterone and 14.7% DHEA-S elevations at 8 weeks in overweight men 40–70 using Shoden extract (21 mg withanolide glycosides/day), but no significant between-group changes in fatigue, vigour, or psychological wellbeing Lopresti et al. 2019. Caveat: most published trials are Indian, with three sponsoring brands (KSM-66, Sensoril, Shoden) overrepresented and small samples (n=40–120 typical).

Rhodiola — moderate, fatigue-focused. Ishaque's 2012 BMC systematic review identified 11 trials; 2/6 physical-fatigue trials and 3/5 mental-fatigue trials reported benefit, with all included studies showing high or unclear risk of bias Ishaque et al. 2012. The strongest single RCT is Olsson 2009: 576 mg/day SHR-5 for 28 days in stress-related fatigue, with the Pines burnout scale improving (p=0.047), two of five Conners' CPT attention indices improving, and morning cortisol response attenuated Olsson et al. 2009. Kasper & Dienel 2017 ran a 12-week multicenter open-label trial (n=118) in burnout patients showing reductions in BOSS I/II burnout, depression, and anxiety scores; open-label design limits interpretation Kasper & Dienel 2017. Compared to ashwagandha, rhodiola has fewer trials, smaller samples, more methodological flaws, and no comparable meta-analysis with a clean point estimate.

Lion's mane — thinnest of the three for cognition. The foundational human trial is Mori 2009: 50–80-year-old Japanese adults with mild cognitive impairment, 3 g/day fruiting-body powder for 16 weeks, significant improvement on the Revised Hasegawa Dementia Scale at weeks 8, 12, 16 — but gains regressed within 4 weeks of stopping Mori et al. 2009. Li 2020 (n=49 mild AD patients, 49 weeks, erinacine-A-enriched mycelia 1.05 g/day) found significant improvements on MMSE and Instrumental Activities of Daily Living versus placebo Li et al. 2020. In healthy young adults, Docherty 2023 (n=41, 28 days, 1.8 g/day) found only an acute single-dose improvement on Stroop reaction time (p=0.005 at 60 min) and a trend-level reduction in subjective stress at 28 days (p=0.051); most chronic cognitive measures showed no significant change Docherty et al. 2023. Mechanism is well-characterised in vitro; human evidence in non-impaired populations is preliminary.

Protocol

Standard clinical doses, by botanical and extract:

• Ashwagandha (KSM-66, root-only, ≥5% withanolides): 600 mg/day, split as 300 mg twice daily with food, for ≥8 weeks before judging effect. Evening dosing favoured when sleep is the primary goal. Sensoril (root+leaf, ≥10% withanolides): 250 mg/day. Shoden (root+leaf, ≥35% withanolide glycosides): 120 mg/day. Doses are not interchangeable across extract types — 600 mg of unstandardised powder ≠ 600 mg KSM-66 in withanolide content NIH ODS Ashwagandha fact sheet.
• Rhodiola (SHR-5 or equivalent, standardised to ≥3% rosavins / ≥1% salidroside): 200–600 mg/day, morning or split AM/early afternoon, taken before mid-afternoon to avoid sleep disruption. Studies range 144–680 mg/day; the well-replicated band is 200–400 mg Olsson et al. 2009.
• Lion's mane: 1–3 g/day of fruiting-body powder, or ~1 g/day of erinacine-A-enriched mycelial extract. Effects in cognitively-impaired populations require sustained intake (≥8 weeks) and reverse on cessation.

All three have a slow-onset profile. Two weeks of dosing is rarely enough to evaluate. The clinical literature converges on 8-week minimum trials for ashwagandha and rhodiola; 12–16 weeks for lion's mane.

Contraindications

Ashwagandha. Avoid in pregnancy (traditional Ayurvedic abortifacient; insufficient safety data). Avoid in clinically apparent thyroid disease or while taking levothyroxine — ashwagandha raises T4 and total/free T3 and can precipitate iatrogenic thyrotoxicosis in supplemented patients NIH ODS Ashwagandha fact sheet. Avoid in active autoimmune disease (immunostimulant action may worsen Hashimoto's, lupus, rheumatoid arthritis). Caution with sedatives, benzodiazepines, immunosuppressants, and antihypertensives. Hepatotoxicity: NIH LiverTox lists ashwagandha as a confirmed cause of clinically apparent liver injury, with onset 2–12 weeks post-initiation, typically cholestatic or mixed pattern, generally idiosyncratic and reversible on discontinuation; rare cases have required transplant or proved fatal in patients with pre-existing cirrhosis NIH LiverTox: Ashwagandha. The UK Food Standards Agency referred ashwagandha to the Committee on Toxicity in 2023 specifically on hepatotoxicity, thyroid, and hypoglycaemia signals.

Rhodiola. Avoid in bipolar disorder — case reports describe manic episodes precipitated by rhodiola, consistent with its monoaminergic mechanism Bredie et al. 2022. Theoretical serotonin syndrome risk when combined with SSRIs, SNRIs, or MAOIs; tramadol and triptans similarly. Avoid evening dosing — the stimulant-like single-dose effect commonly causes insomnia. Caution in patients with anxiety disorders involving agitation (rhodiola can worsen jitteriness at higher doses).

Lion's mane. Generally well-tolerated; rare reports of contact dermatitis and gastrointestinal upset. In the Li 2020 trial, 4/49 subjects dropped out for abdominal discomfort, nausea, or rash Li et al. 2020. Theoretical concern about anticoagulant interaction; insufficient pregnancy data.

Misconceptions

(1) "Adaptogen" is a clinical pharmacology category — it is not, in the Western regulatory sense; it is a tradition-grounded heuristic that has accumulated some mechanistic legitimacy through Panossian-school work on HPA-axis and stress-response pathways but no FDA/EMA pharmacological-class recognition. (2) Doses are interchangeable across extracts — they are not (see §3b/Protocol). (3) Ashwagandha is "natural and therefore safe" — the LiverTox listing and the FSA referral disprove this; injury is rare but real and idiosyncratic NIH LiverTox: Ashwagandha. (4) Lion's mane "boosts" memory in healthy young adults — current human evidence shows only acute single-dose effects and a trend on stress at 28 days; chronic cognitive enhancement in healthy youth is unproven Docherty et al. 2023. (5) Adaptogens work in days — clinical effect requires ≥8 weeks for the HPA-axis-mediated outcomes; reports of immediate effect are likely placebo or single-dose sympathomimetic action.

Failure modes

The five most common reasons "I tried adaptogens and they didn't work": (1) wrong extract or dose — unstandardised bulk powder with unknown withanolide/rosavin content; (2) too short a trial — quitting at 2–4 weeks before HPA-axis-mediated effects manifest; (3) wrong botanical for the goal — taking rhodiola for sleep (it is stimulant-leaning) or ashwagandha for acute focus (it is sedative-leaning); (4) stacked products that mask which compound is doing what; (5) baseline cortisol or sleep is already normal, leaving little room to move. Industry-funded trials enrich for stressed populations, where effects are largest — translating to a normal-stress baseline halves the expected effect size.

Practicalities

Cost at standard dosing: ashwagandha KSM-66 at 600 mg/day is roughly $20–60/year; rhodiola SHR-5-grade extract roughly $80–200/year; lion's mane (concentrated mycelial extract or organic fruiting-body powder) $60–250/year. Third-party-tested brands (NSF, USP, Informed Sport) cost more but mitigate the heavy-metal and adulteration risk endemic to herbal supplements imported from India and China. Effort is trivial — one to two capsules daily, no special timing constraints beyond AM-leaning for rhodiola and PM-leaning for ashwagandha-for-sleep. Trial-and-error is the failure mode: most users cycle through several brands and botanicals before finding what works, which inflates effective cost.

History

The adaptogen concept originated in 1947 in Soviet wartime pharmacology (Lazarev, then Brekhman and Dardymov in the 1960s), aimed at finding compounds to improve soldier and worker performance under physiological stress without amphetamine-like dependence. Soviet pharmacological reviews catalogued the property in Eleutherococcus senticosus, Schisandra chinensis, and Rhodiola rosea. Ashwagandha entered Western use via the Ayurvedic tradition (≥3,000-year history as a rasayana — rejuvenative tonic). Lion's mane comes from East Asian culinary and traditional Chinese / Japanese medical use (shishigashira, yamabushitake). The convergence of three culturally distinct herbal traditions under a single Western marketing label (~2010–2020) is recent; the pharmacological case for grouping rhodiola and ashwagandha is reasonable (both modulate HPA-axis); for lion's mane it is much weaker (the mechanism is neurotrophin synthesis, not stress-response modulation).

Alternatives

For stress / cortisol reduction with stronger evidence: cognitive-behavioural therapy for stress, mindfulness-based stress reduction, regular aerobic exercise (effect sizes for chronic stress comparable or larger). For sleep, sleep-hygiene interventions, CBT-I, and (where appropriate) low-dose melatonin or magnesium glycinate have larger evidence bases than ashwagandha. For energy/fatigue, addressing the medical cause (iron deficiency, sleep apnoea, hypothyroidism, depression) precedes any adaptogen trial. For mild-to-moderate depression, SSRIs and exercise have RCT bases orders of magnitude larger than rhodiola. For cognitive enhancement in healthy adults, sleep, aerobic exercise, and caffeine have stronger evidence than lion's mane.

Stakes

Skipping adaptogens entirely is not a meaningful health risk — none of the three is essential, and the absolute effect sizes are modest. The reader who tries them and notices nothing has lost $20–250 and 8 weeks; the reader who skips them entirely is missing what is, at best, a moderate adjunctive intervention for chronic stress and sleep difficulty.

Payoff

Best-case projection — ashwagandha at 600 mg/day for 12 weeks in an adult with chronically elevated stress and sub-clinical insomnia: ~30% drop in serum cortisol, modest PSS reduction (4–5 points), 30–60-minute earlier sleep onset, and subjectively-felt resilience to daily stressors — onset around weeks 3–4, plateau around weeks 8–12 Cheah et al. 2021 Bachour et al. 2025. Rhodiola best-case in a burnout-presentation worker: reduced mental fatigue by week 2, improved morning cortisol response, sustained at 12 weeks Olsson et al. 2009. Lion's mane best-case in mild cognitive impairment: improved cognitive screening scores by week 16, reversing on discontinuation Mori et al. 2009. None of these is transformative; all are modest adjunctive shifts.

The credibility range

Optimist case

The ashwagandha sleep + cortisol meta-analyses (Cheah 2021; Bachour 2025) cover >1,200 participants across >15 RCTs, with consistent direction of effect (cortisol down, sleep up, anxiety down), moderate-quality evidence, dose-response signal (600 mg outperforms 250 mg), and a plausible multi-pathway mechanism (HPA-axis dampening + GABAergic action + glucocorticoid-receptor resensitisation). The signal has now been replicated outside India in Western samples. Rhodiola's burnout/fatigue evidence, though weaker, lines up with a coherent mechanism (mild MAO-inhibition + HSF1-pathway stress response) and decades of Eastern European use without dependence or serious adverse events. Lion's mane has the most-defined molecular mechanism of any nootropic candidate, with two distinct NGF-inducing diterpenoid families that cross the blood-brain barrier, plus two positive RCTs in cognitively impaired populations. The "adaptogen" frame, while marketing-heavy, points at a real pharmacological pattern: multi-target compounds that modulate the stress system without producing dependence — and several work in a class where most modern psychiatric tools (benzodiazepines, stimulants) do produce dependence.

Skeptic case

(i) Trial-quality is poor. Most ashwagandha RCTs are small (n<100), short (8–12 weeks), Indian, and funded by KSM-66 / Sensoril / Shoden manufacturers. Risk-of-bias concentrates in lack of allocation concealment and incomplete outcome reporting; the rhodiola literature is worse. (ii) Heterogeneity is high (I²=50–64% in the sleep meta-analysis), suggesting the pooled effect mixes meaningful responders with non-responders. (iii) The "non-specific resistance" definition is unfalsifiable — adaptogens are claimed to lower cortisol when high and raise it when low, which sidesteps the requirement that a drug have a directional effect. (iv) Lion's mane in healthy young adults shows mostly null chronic results (Docherty 2023); the MCI/AD effects do not generalise. (v) Hepatotoxicity is no longer "theoretical" — the LiverTox listing and the UK FSA referral place ashwagandha in the same idiosyncratic-injury class as kava and green-tea extract. (vi) The "felt effect" is heavily placebo-amplified: open-label ashwagandha and rhodiola users report transformation; rigorously-blinded RCTs report 0.4–0.6 SD point estimates that translate to modest, not dramatic, clinical change. (vii) Lopresti 2019 found endocrine changes but no felt-experience benefit — the kind of small-but-significant lab result the literature rewards while users see nothing.

Author's call

Ashwagandha at 600 mg/day for ≥8 weeks is the best-evidenced single adaptogen, with replicated, moderate-quality effects on cortisol, perceived stress, sleep onset, and sleep quality — large enough to be worth trying, small enough that it should not be expected to replace CBT, exercise, or sleep-hygiene interventions. Rhodiola is a defensible second choice for stress-related fatigue or burnout presentation, with weaker but coherent evidence. Lion's mane is a worth-watching candidate for cognitive aging, but the human evidence outside MCI/early AD is too thin to recommend confidently in healthy adults. The category as marketed — interchangeable, indication-agnostic, fast-acting — is oversold. The category as actually studied — botanical-specific, dose-specific, slow-onset, modestly effective — has a real but moderate role.

Stakeholder and incentive map

• Commercial: KSM-66 (Ixoreal Biomed), Sensoril (Natreon), Shoden (Arjuna Natural) — Indian botanical-extract firms that fund and design most ashwagandha RCTs. Major US/EU supplement retailers (Thorne, Pure Encapsulations, Gaia Herbs, NOW Foods) drive distribution. Adaptogen-branded functional-beverage startups (Recess, Kin Euphorics, Four Sigmatic) are a fast-growing channel.
• Professional: Ayurvedic practitioners (ashwagandha) and naturopathic/integrative-medicine clinicians (all three) — the credibility cohort. Mainstream psychiatry/sleep medicine remains skeptical; few clinical guidelines mention adaptogens.
• Cultural: Andrew Huberman, Tim Ferriss, Joe Rogan — podcast-driven adoption has been the dominant Western adoption vector since ~2018. Wellness Instagram and TikTok amplify single-anecdote reports.
• Skeptic / counter-incentive: NIH LiverTox, UK FSA / COT, ConsumerLab and Labdoor (which routinely find label-claim violations and contamination), academic psychopharmacology reviewers. Mainstream sleep and anxiety guidelines (AASM, NICE) do not include adaptogens.

Population variability

• Ashwagandha: largest effects in adults with elevated baseline PSS or diagnosed insomnia; small or null effects in low-stress healthy adults. Most trial samples are Indian; the few Western replications are consistent in direction. Male testosterone signal is observed mostly in fatigued/overweight 40–70-year-old men Lopresti et al. 2019.
• Rhodiola: most-studied in burnout-presentation working adults; Eastern European samples dominate. Bipolar-spectrum individuals are at risk of mania Bredie et al. 2022.
• Lion's mane: positive cognitive evidence concentrated in 50–80-year-old subjects with MCI or mild AD Mori et al. 2009 Li et al. 2020; healthy young-adult evidence largely null on chronic measures Docherty et al. 2023.
• Across all three: pregnancy, breastfeeding, autoimmune disease, hepatic impairment, bipolar disorder, and concomitant psychiatric medication change the calculus substantially.

Knowledge gaps

• No large (n>500), independently-funded, multicentre Western RCT exists for any of the three botanicals.
• Long-term safety beyond 12 months is sparsely studied; the Salve 2025 12-month KSM-66 safety study (n=191) is the longest published.
• Hepatotoxicity incidence per million daily doses of ashwagandha is unknown — current case series cannot distinguish a 1-in-10⁴ from 1-in-10⁶ rate.
• Head-to-head trials comparing extract types (KSM-66 vs Sensoril vs Shoden for ashwagandha; SHR-5 vs other rosavin/salidroside ratios for rhodiola) are absent.
• The "raises cortisol when low / lowers when high" amphoteric claim has not been formally tested — all positive trials enrol elevated-cortisol populations.
• Lion's mane evidence in healthy adults (the primary marketed use) is two small pilots; the chronic-effect signal needs replication at n>200.
• Interaction studies with SSRIs, levothyroxine, and immunosuppressants are limited to case reports and theoretical extrapolation.

Scope and the brief. The brief named ashwagandha, rhodiola, and lion's mane as the three botanicals; named consequences were stress response, mood, sleep, cognition, energy, and hormone regulation. The article covers all six, though hormone regulation is mostly discussed under ashwagandha-specific effects (testosterone in aging men via Lopresti 2019) rather than as a separate addressing section — the felt-experience case for hormone effects is weak (Lopresti found significant lab changes but no felt-vitality benefit), so it sits in the evidence section rather than warranting its own.

Why grouped under one entry. Consumer-supplement market reality: these three are sold, stacked, and recommended interchangeably under the "adaptogen" banner. Per entry.md §1a, "substance plus all meaningful consequences" — and here the substance is "the category of daily adaptogenic-botanical supplementation," with three named members. Splitting into three separate entries would duplicate the mechanism, contraindications, and protocol sections and lose the comparative framing the reader actually needs ("which of these is for my problem"). If lion's mane develops a much larger young-adult evidence base, splitting it off into a standalone nootropic entry is defensible.

Rating difficulties. evidence: 3 was the hardest call. Ashwagandha alone arguably justifies a 4 — two meta-analyses, dose-response, replicated effect. But the entry covers all three, and rhodiola's literature has unclear/high risk of bias across the board (Ishaque 2012), while lion's mane evidence in healthy adults is essentially one pilot (Docherty 2023). A "3" reflects the weighted average across the category as the reader will actually encounter it. controversy: 3 reflects the LiverTox listing + UK FSA referral on ashwagandha plus the active dispute about whether "adaptogen" is a meaningful category at all.

Hard exclusions. Ginseng, eleuthero, holy basil, schisandra, cordyceps, reishi, maca, bacopa — all marketed as adaptogens, all left out. Adding them would have pushed the entry past the point of being useful as a comparative-framework article and into encyclopedia territory. Listed as forward pointers in out-of-scope. Withaferin-A anticancer preclinical work was excluded as out-of-scope for a daily-supplement entry.

Separate-entry candidates. If the catalogue scales, Lion's Mane for Cognitive Aging deserves its own entry — it's mechanistically distinct from the HPA-axis adaptogens, the target population is different (60+ with mild decline, not stressed 30-somethings), and the evidence framing is "promising therapeutic for early dementia" rather than "wellness supplement." Similarly, Hericium erinaceus in Mild Cognitive Impairment may belong in the medical category as a clinician-discussed adjunct rather than under supplements.

Future-link candidates. Once written, this entry should cross-link to: Cortisol Awakening Response, HPA-axis dysregulation, CBT-I, Magnesium glycinate, L-theanine, and any entry on chronic-stress management.

Honest hedge on the optimist case. The skeptic case is stronger than the article's tone lets on. The Indian-trial concentration, brand-funding pattern, and small sample sizes are red flags that an academic reviewer would weight more heavily than the article does. The author's call landed pro-ashwagandha-for-sleep because the meta-analysis effect is consistent and clinically meaningful, but a careful reader of the underlying trials would notice the same flaws this note names. The hepatotoxicity warning in contraindications + the misconceptions section are the article's main hedges against under-stating risk.

Audience scoping not applied. The substance is general; the Lopresti testosterone signal in older men is a sub-finding rather than the dominant frame, so no audience field was set. If the entry were split, a lion's-mane-cognitive-aging spin-off would scope to ages: ["60+"].