evipedia.ai
← Back

Ashwagandha for Health & Longevity

Evidence Review created on 09/02/2026 using AI4L / Opus 5

Also known as: Withania somnifera, Indian Ginseng, Winter Cherry, Asgandh, Asgand, Poison Gooseberry

Motivation

Ashwagandha (Withania somnifera) is a small shrub whose root has been used in Indian traditional medicine for roughly three thousand years. Its extracts are rich in steroidal compounds called withanolides, which appear to soften the hormonal chain running from the brain to the adrenal glands — the pathway governing how strongly the body reacts to stress. That single property is why a herb from a classical medical system now sits in mainstream supplement aisles.

Over the past two decades ashwagandha has become one of the best-selling botanical extracts in North America and Europe. Its rise has been accompanied by a rapid accumulation of short placebo-controlled human studies, most of them paid for by the companies selling the extracts, and by a smaller but growing set of medical reports describing liver damage in people taking it.

This review examines what that body of work supports for readers who treat sleep, stress load and physical capacity as levers on long-term health: how large and durable the reported effects are, what the hormonal changes amount to, how serious the liver and thyroid signals are, and what the literature describes on dosing, product quality and monitoring.

Benefits - Risks - Protocol - Conclusion

High-level overviews of ashwagandha from independent experts and longevity-focused publications.

  • How to Control Your Cortisol & Overcome Burnout - Andrew Huberman

    A protocol-level treatment of the cortisol rhythm ashwagandha acts on, with a dedicated supplement segment naming the herb and citing the stress trial that anchors its evidence base.

  • Ashwagandha root extract boosts stress resistance. - FoundMyFitness

    A compact walkthrough of the 64-person chronic-stress trial that anchors most ashwagandha marketing, covering the cortisol and mood-score changes and how adaptogenic plants are thought to act.

  • Ashwagandha: Benefits, Uses, and Side Effects - Stephen Rose

    A longevity-oriented primer on withanolide chemistry and on the stress, cognition and immune claims, written for readers assessing the herb as a healthspan tool rather than a treatment.

  • Ashwagandha’s Brain Benefits - Rick Wilson

    Assembles the neurological case — cognitive performance in healthy adults, calming-neurotransmitter activity, and amyloid and stroke findings — and is explicit about which claims rest only on preclinical work.

  • RHR: From Wired & Tired to Calm & Clear: My Top Nutrients for Mood, Focus, and Sleep - Chris Kresser

    Devotes a segment to ashwagandha among five calming nutrients, setting the cortisol and sleep trial data alongside clinical experience of gastrointestinal tolerability problems and combination use.

One of the six priority platforms is not represented above. Peter Attia’s ashwagandha material sits inside subscriber-only Ask-Me-Anything show notes and audio that cannot be read without a membership, and the section is capped at five items with no more than one per source.

Grokipedia

  • Ashwagandha

    The site’s primary article on the intervention, covering withanolide chemistry, pharmacokinetics, the modern clinical literature, dosage practice, reported safety concerns and regulatory status in one place.

Examine

  • Ashwagandha

    Grades ashwagandha’s evidence outcome by outcome across 29 trials and 8 meta-analyses, and carries a separate safety database covering side effects, drug interactions and pregnancy status.

ConsumerLab

  • Ashwagandha Supplements Reviews & Top Picks

    Reports independent laboratory measurement of withanolide content and heavy-metal contamination across commercial ashwagandha products, and names which ones matched their labels closely enough to be approved.

Systematic Reviews

The strongest systematic reviews and meta-analyses of ashwagandha in humans are listed below, with the caveat that most of the underlying trials were funded by the extract manufacturers themselves — principally Ixoreal Biomed (KSM-66), Natreon (Sensoril) and Arjuna Natural (Shoden) — parties with a direct financial interest in a favourable result.

Mechanism of Action

Ashwagandha root is a mixture, and its activity is attributed mainly to withanolides — steroidal lactones including withaferin A, withanolide A and 12-deoxywithastramonolide — plus withanosides and sitoindosides. The dominant explanation is dampening of the HPA axis (hypothalamic-pituitary-adrenal axis, the brain-to-adrenal-gland circuit that releases cortisol under stress). Pooled trial data show cortisol falling substantially with supplementation, which is consistent with this account (Fornalik et al., 2026).

A competing, more direct explanation is receptor-level sedation: withanolides act on GABA-A receptors (gamma-aminobutyric acid receptors, the brain’s main calming switch), which would produce anxiety and sleep effects without any hormonal step. Withanolides also inhibit NF-κB (nuclear factor kappa B, a master switch for inflammatory genes), activate Nrf2 (nuclear factor erythroid 2-related factor 2, which turns on antioxidant defences) and inhibit acetylcholinesterase (the enzyme that clears the memory neurotransmitter acetylcholine). Whether the cortisol drop causes the clinical effects or merely accompanies them remains unresolved (D’Cruz & Andrade, 2022).

Human pharmacokinetic data — how the body absorbs, distributes and clears withanolides — are thin. Withanolides are fat-soluble, absorbed within one to two hours, cross the blood-brain barrier, and are cleared with a short plasma half-life of a few hours; the clinical effect nonetheless builds over weeks, implying adaptation rather than acute drug action. In laboratory systems withanolides inhibit CYP3A4 and CYP2B6 (liver enzymes that metabolise many medicines) and interact with P-glycoprotein (a cellular drug-export pump).

Historical Context & Evolution

Ashwagandha entered the written record as a rasayana — a rejuvenating tonic — in the classical Ayurvedic compendia. Its Sanskrit name means “smell of the horse”, read both as a description of the root’s odour and as a promise of a stallion’s vigour. Traditional use was broad: emaciation and wasting states, weakness in the elderly, poor sleep, and male sexual complaints. The root, not the leaf, was the traditional preparation, usually taken with milk or clarified butter.

Systematic pharmacological work began in Indian laboratories in the 1950s and 1960s, which documented sedative, anti-inflammatory and anti-tumour activity in animals. In the 1960s Soviet pharmacology introduced the term “adaptogen” for plants that raise non-specific resistance to stress, and ashwagandha was absorbed into that category. Commercial standardized extracts arrived in the 2000s and 2010s, and a 2012 placebo-controlled trial reporting a large drop in perceived stress and serum cortisol (Chandrasekhar et al., 2012) became the anchor citation for a marketing wave.

The direction of travel then changed. From 2019 onward, hepatology groups in Iceland, the United States, India, Poland and Spain published causality-assessed cases of cholestatic liver injury (a blockage of bile flow out of the liver, causing jaundice and itching) (Björnsson et al., 2020), and a Danish national risk assessment prompted restrictions on ashwagandha in food supplements on hormonal grounds — a reading that later analyses have contested (Vollmer & Brendler, 2026). Neither the efficacy nor the safety literature has settled.

Expected Benefits

The trials underlying the graded items below were, in the large majority, sponsored by ashwagandha extract manufacturers — Ixoreal Biomed, Natreon and Arjuna Natural — who profit directly from positive findings; this is reflected in the grading and again in the Conclusion.

High 🟩 🟩 🟩

Reduced Stress and Anxiety

The most replicated effect. Pooled across 12 randomized controlled trials in 1,002 adults, ashwagandha lowered anxiety and perceived stress on validated instruments by large standardized margins, and a separate pooled analysis found a mean 5.96-point drop on the Hamilton Anxiety Rating Scale (Akhgarjand et al., 2022; Fatima et al., 2024). Serum cortisol falls in parallel. Heterogeneity between trials is very high, most are eight weeks or shorter, and the pooled certainty was rated low — so the direction is secure while the size is not.

Magnitude: Standardized mean difference (an effect size expressed in pooled standard-deviation units) of −1.55 for anxiety and −1.75 for perceived stress versus placebo; equivalently, a mean reduction of 5.96 points on the Hamilton Anxiety Rating Scale.

Improved Sleep Quality

Ashwagandha shortens the time taken to fall asleep and raises sleep efficiency, with the effect concentrated in people who meet criteria for insomnia rather than in good sleepers. The pooled estimate comes from five randomized trials in 400 adults using the Pittsburgh Sleep Quality Index and actigraphy (Cheah et al., 2021; Langade et al., 2021). Benefit was clearest at 600 mg daily or above and at eight weeks or longer. Whether the sleep gain is separate from the anxiety gain has not been disentangled.

Magnitude: Standardized mean difference −0.59 for overall sleep versus placebo, with the largest subgroup effects in diagnosed insomnia at doses of 600 mg per day or more.

Increased Muscle Strength and Size with Resistance Training

In trained and training adults, ashwagandha added to a resistance programme increases maximal strength and limb circumference beyond the programme alone, and blunts the creatine kinase rise that marks exercise-induced muscle damage (Wankhede et al., 2015). A 2026 meta-analysis of 20 trials confirmed the strength signal and showed it is restricted to physically active people, with no effect in untrained or clinical groups (Zhu et al., 2026). Body weight and body fat percentage were unchanged in pooled analysis.

Magnitude: Standardized mean difference 0.58 for muscle strength overall and 1.03 in physically active participants; in the anchor trial, bench-press one-repetition maximum rose 46.0 kg versus 26.4 kg on placebo over eight weeks.

Improved Memory and Cognitive Performance

Pooled across 20 randomized trials in 1,249 adults, ashwagandha improved memory, attention and processing speed, executive function and visuospatial ability on standard neuropsychological batteries (Zhu et al., 2026). The largest single contribution comes from an eight-week trial in adults with MCI (mild cognitive impairment, measurable memory loss that falls short of dementia), where immediate and general memory, sustained attention and information-processing speed all improved (Choudhary et al., 2017). Populations were heterogeneous and no trial has run past twelve weeks.

Magnitude: Standardized mean difference 0.52 for memory, 0.29 for attention and processing speed, and 0.42 for executive function versus placebo.

Improved Cardiorespiratory Fitness

Ashwagandha raises VO₂max (maximal oxygen uptake, the ceiling on aerobic capacity and one of the strongest single predictors of all-cause mortality) in both athletes and untrained adults. The pooled estimate comes from four randomized controlled trials in 142 participants (Pérez-Gómez et al., 2020), and a subsequent three-level meta-analysis of exercise-performance outcomes found the same direction across endurance measures (Li et al., 2026). Heterogeneity is high and the pooled confidence interval sits close to zero, so the estimate is imprecise rather than contested.

Magnitude: Mean difference of 3.00 mL/kg/min in VO₂max versus placebo (95% CI 0.18 to 5.82; CI meaning confidence interval, the range within which the true value most likely lies).

Medium 🟩 🟩

Higher Serum Testosterone in Men

Two 2026 meta-analyses agree that ashwagandha raises circulating testosterone in men and not in women, and an earlier systematic review of herbal testosterone agents singled out ashwagandha root as one of only two ingredients with a positive signal (Fornalik et al., 2026; Smith et al., 2021). Dehydroepiandrosterone sulphate rises in parallel in overweight middle-aged men (Lopresti et al., 2019). This is graded Medium because the endpoint is a hormone concentration rather than a symptom or function outcome, and no trial has linked the rise to a clinical result.

Magnitude: Mean difference of 57.4 ng/dL in men and 5.1 ng/dL in women (not significant); standardized mean difference 0.33 across pooled trials.

Relief of Perimenopausal Symptoms

In perimenopausal women, eight weeks of 600 mg daily reduced total Menopause Rating Scale scores across psychological, physical and urogenital domains and improved menopause-specific quality of life, alongside a rise in estradiol and falls in follicle-stimulating and luteinizing hormone (Gopal et al., 2021). A smaller 2025 replication in 60 women reported the same direction (Vani et al., 2025). The evidence rests on one adequately sized manufacturer-funded trial plus one small confirmation, so the effect is credible but not yet independently established.

Magnitude: Statistically significant reduction in total Menopause Rating Scale score versus placebo across all three symptom domains over eight weeks; the trial reports domain-level significance rather than a pooled point estimate.

Reduced Body Weight and Food Cravings Under Chronic Stress

In 52 adults under chronic stress, 600 mg daily for eight weeks reduced perceived stress, food-craving scores, serum cortisol, body weight and body mass index relative to placebo (Choudhary et al., 2017). The mechanism proposed is indirect — lower cortisol reducing stress-driven eating rather than any metabolic action. Pooled analysis of the wider trial set found no effect on body weight or body fat in unselected adults (Zhu et al., 2026), so the benefit appears confined to people whose weight gain is stress-linked.

Magnitude: Significant reductions in body weight, body mass index and food-craving scores versus placebo over eight weeks in stressed adults; the trial reports significance without publishing a between-group kilogram difference.

Low 🟩

Normalization of Thyroid Indices in Subclinical Hypothyroidism ⚠️ Conflicted

One 50-person pilot trial in adults with raised thyroid-stimulating hormone reported normalized thyroid indices at 600 mg daily (Sharma et al., 2018); pooled trial data show no such change (Fornalik et al., 2026). Net reading: a small effect visible only when thyroid function is already borderline.

Magnitude: Mean thyroxine increase of 0.61 µg/dL in pooled trials, with no significant change in thyroid-stimulating hormone or triiodothyronine.

Improved Male Sexual Function and Semen Parameters

Ashwagandha improved self-reported sexual performance scores in a randomized trial of men (Chauhan et al., 2022), and a systematic review reported higher sperm concentration and motility in men with low counts (Nasimi Doost Azgomi et al., 2018). The underlying trials are small, single-site and use non-standardized questionnaires.

Magnitude: In the dedicated trial the Derogatis Interview for Sexual Functioning–Male total score rose 9.8 points further than on placebo (95% CI 8.87 to 10.73) over eight weeks; the literature reports no pooled figure for sperm concentration.

Lowered Fasting Glucose and Lipids ⚠️ Conflicted

Five clinical studies in diabetes show favourable shifts in fasting glucose, glycated haemoglobin and lipids, though the pooled review rates them too sparse to be robust (Durg et al., 2020); a 12-month cohort in healthy adults found no change (Salve et al., 2026). Net reading: modest and baseline-dependent.

Magnitude: Direction is downward for fasting glucose and glycated haemoglobin where these are already raised, with no change at a normal baseline; the only pooled figure the literature reports comes from animal rather than human data.

Reduced Knee Joint Pain and Stiffness ⚠️ Conflicted

A 12-week placebo-controlled trial in 60 adults with knee joint pain reported dose-dependent reductions in pain, stiffness and disability scores (Ramakanth et al., 2016); a network meta-analysis of botanical extracts confirmed only the stiffness component (Yan et al., 2025). Net reading: a stiffness effect, not an analgesic one.

Magnitude: Mean difference of 4.16 points in pooled knee-osteoarthritis stiffness score versus placebo (95% CI 2.43 to 5.90); the single dedicated trial reports significant symptom reduction at 250 mg twice daily without publishing a between-group figure.

Speculative 🟨

Geroprotective and Healthspan Effects

Root extract extends lifespan in Caenorhabditis elegans; one proposed mechanism is activation of telomerase (the enzyme that maintains chromosome ends). No human longevity data exist — the basis is model-organism work only (Vittal & Vinciguerra, 2025).

Anticancer and Chemotherapy-Adjunct Activity

Withaferin A kills tumour cells in culture and shrinks implanted tumours in animals through heat-shock pathways. Human evidence is limited to early-phase trials; the basis remains laboratory and animal work (Tewari et al., 2022).

Immune Cell Activation

A randomized trial reported higher antibody levels and lymphocyte counts in healthy adults after 30 days. No infection or illness outcome was measured, so the basis is immune biomarkers only (Tharakan et al., 2021).

Benefit-Modifying Factors

  • Baseline stress and sleep quality: The effect scales inversely with baseline function. Pooled sleep data show benefit concentrated in people meeting insomnia criteria, and anxiety effects are largest in those with the highest entry scores (Cheah et al., 2021); well-rested, low-stress individuals should expect little.

  • Baseline testosterone and training status: Strength gains appear only in physically active people, not the untrained (Zhu et al., 2026). Testosterone rises are largest in overweight, fatigued middle-aged men with lower starting values, and are absent in women.

  • Withanolide dose and extract standardization: Response tracks withanolide delivery, not milligrams of powder. Trials using 5% withanolide root extract at 600 mg daily are the reference; products delivering under 6 mg of withanolides per serving have no matching trial evidence.

  • Sex: Testosterone and dehydroepiandrosterone sulphate responses are male-specific in pooled analysis (Fornalik et al., 2026). Women show the stress, sleep and cognitive effects but not the androgen effects, and the perimenopausal benefit is a separate, female-specific finding on sex-steroid balance.

  • Pre-existing thyroid status: Thyroid indices move measurably only when they are already abnormal. Adults with borderline-high thyroid-stimulating hormone see normalization; adults with normal thyroid function show no change, which cuts both ways as benefit and risk.

  • Age: Cognitive and quality-of-life gains are largest at the older end of the target range — the 12-month observational safety cohort recorded global improvement most often in participants aged 50 and over (Salve et al., 2026), while strength gains were documented in younger training populations.

  • Genetic polymorphisms: No pharmacogenetic variant has been validated for ashwagandha response. Because withanolides interact with CYP3A4 and CYP2B6 (drug-metabolising liver enzymes), variation in those genes is a plausible but untested source of differing exposure and effect.

Potential Risks & Side Effects

High 🟥 🟥 🟥

Gastrointestinal Upset

The most frequent complaint across the trial literature: nausea, loose stools, abdominal cramping, bloating and reflux, typically in the first two weeks and typically dose-related. All recorded events in the 12-month observational safety cohort were mild and resolved without intervention (Salve et al., 2026), and a randomized safety trial at 600 mg daily recorded none in either arm (Verma et al., 2021). Taking the extract with food and splitting the dose reduces it. People with existing gastrointestinal conditions report it more often.

Magnitude: In the 12-month cohort of 191 adults, 18 mild adverse events in total were recorded, gastrointestinal complaints predominating; short trials report rates comparable to placebo.

Daytime Drowsiness and Sedation

Ashwagandha’s action on calming neurotransmitter receptors produces sedation that does not always stay confined to bedtime. Drowsiness is the second most consistently reported adverse effect across randomized trials and is listed as a primary side effect in independent safety databases (Fatima et al., 2024). It is dose-related and reversible, matters most for driving and machine operation, and compounds with alcohol, antihistamines and prescription sedatives. Evening-only dosing is the usual mitigation.

Magnitude: Direction is a consistent dose-related increase in reported drowsiness relative to placebo; trials report it as a listed adverse event without publishing an incidence figure.

Medium 🟥 🟥

Herb-Induced Liver Injury

The most serious documented harm. Independent hepatology groups in Iceland, the United States, India, Poland, Spain and Bosnia have published causality-assessed cases of cholestatic or mixed liver injury with jaundice and severe itching, appearing two to twelve weeks after starting (Björnsson et al., 2020; Philips et al., 2023). Most cases resolve in one to five months; in patients with pre-existing chronic liver disease the Indian series recorded liver failure and death. Graded Medium because the signal comes from consistent observational case series, not from controlled trials.

Magnitude: Latency 2–12 weeks; jaundice and itching persisting 5–20 weeks; liver tests normalize within 1–5 months in most cases, but 3 of 8 single-ingredient Indian cases with underlying chronic liver disease progressed to liver failure and died. No population incidence rate has been established.

Thyroid Hormone Elevation and Thyrotoxicosis ⚠️ Conflicted

Ashwagandha raises thyroxine modestly in pooled trial data and normalized thyroid indices in adults with subclinical hypothyroidism, and isolated case reports describe thyrotoxicosis (an over-active thyroid state with palpitations, tremor and weight loss) (Fornalik et al., 2026; Sharma et al., 2018). Against this, a randomized safety trial in healthy adults found no thyroid change (Verma et al., 2021). Net reading: a small genuine thyroid-stimulating effect that is clinically silent in people with normal thyroid function but can tip those already on thyroid replacement or with borderline function.

Magnitude: Pooled thyroxine increase of 0.61 µg/dL with no change in thyroid-stimulating hormone or triiodothyronine; overt thyrotoxicosis is confined to isolated case reports.

Low 🟥

Headache

Headache is named across the randomized trial literature alongside gastrointestinal complaints and drowsiness, and independent supplement-testing summaries list it. The pooled safety review of 23 root-extract trials found it infrequent and present in placebo arms too, so attribution is uncertain (Coope et al., 2026). It is transient.

Magnitude: Direction is a low-frequency occurrence that does not clearly exceed placebo, concentrated in the first weeks; the pooled safety review lists headache among mild events in both arms without publishing an incidence figure.

Loss of Immunosuppressant Effect and Transplant Rejection

Ashwagandha stimulates lymphocyte activity, opposing transplant anti-rejection drugs. A kidney-transplant recipient stable for two years on mycophenolate, tacrolimus and prednisone had biopsy-confirmed rejection two weeks after starting it (Sriperumbuduri et al., 2020). The evidence is one uncontrolled report, but the consequence is irreversible.

Magnitude: One documented case of acute allograft rejection two weeks after initiation; the literature reports no incidence figure because only isolated reports exist.

Under-Dosed and Contaminated Products

Independent pharmacopoeial testing found most commercial products failed quality standards and contained undisclosed leaf and aerial material, which carries higher predicted toxicity than root (Avula et al., 2026). The risk is indirect — an inert product or unintended plant-part exposure rather than a measured clinical outcome.

Magnitude: Of 25 commercial supplements tested, over 44% failed British Pharmacopoeia and 60% failed United States Pharmacopoeia standards; only 10 met both, and only 2 were confirmed root-derived.

Speculative 🟨

Reproductive Toxicity from Non-Root Plant Parts

Computational toxicity modelling predicts higher liver and reproductive toxicity for leaf and aerial-part withanolides than for root, and several implicated products contained leaf. No human reproductive outcome has been measured (Ronen et al., 2025).

Harm in Pregnancy and Lactation

Traditional texts describe pregnancy-ending use, and high-dose animal work is reassuring but limited. No human pregnancy safety data exist, which is why independent safety sources recommend avoidance rather than caution (Ronen et al., 2025).

Risk-Modifying Factors

  • Pre-existing liver disease: The strongest risk modifier. In the Indian case series (Philips et al., 2023), all three deaths occurred in patients with underlying chronic liver disease; in previously healthy patients the same injury was prolonged but self-limiting.

  • Concurrent hepatotoxic (liver-damaging) supplements: Risk stacks. Green tea extract, Garcinia cambogia, turmeric and red yeast rice carry independent liver-injury signals, and several reported ashwagandha cases involved multi-ingredient products or concurrent botanical use.

  • Baseline liver enzymes and bilirubin: A raised alanine aminotransferase or bilirubin before starting removes the ability to detect drug-induced injury against baseline and marks a liver already under load; both argue against initiation.

  • Thyroid status and thyroid medication: Adults on levothyroxine or with borderline thyroid-stimulating hormone are the group in whom the small thyroxine-raising effect becomes clinically visible, potentially producing over-replacement symptoms at an unchanged prescription dose.

  • Sex: Androgen-related effects and their downside — acne, irritability, altered sex-hormone balance — are male-specific. Reported liver-injury case series are male-predominant, though whether that reflects biology or usage patterns is unknown.

  • Age: Older adults take more concurrent medications and have less liver reserve, raising both interaction and liver-injury consequence. Sedation also compounds fall risk in this group, which matters most at the older end of the target range.

  • Genetic polymorphisms: No validated variant predicts ashwagandha liver injury. Human leukocyte antigen haplotypes govern susceptibility for several other herb-induced liver injuries, making them a plausible but currently untested candidate here.

Key Interactions & Contraindications

  • Immunosuppressants (tacrolimus, cyclosporine, mycophenolate, prednisone): Absolute contraindication. Ashwagandha’s immune-stimulating action opposes them and has been followed by acute transplant rejection. No mitigation exists other than avoidance.

  • Thyroid hormone replacement (levothyroxine, liothyronine): Caution. Additive thyroid-hormone effect can produce over-replacement — palpitations, tremor, weight loss. Mitigation: thyroid-stimulating hormone and free thyroxine at six to eight weeks, with the prescription dose adjusted accordingly.

  • Sedatives and hypnotics (benzodiazepines, zolpidem, gabapentin, barbiturates): Caution. Additive central nervous system depression causing excess sedation and impaired coordination. Mitigation: separated dosing, no driving after evening dosing, and initiation at the low end.

  • Antidiabetic drugs (metformin, sulfonylureas, insulin): Monitor. Ashwagandha modestly lowers fasting glucose, which can compound into hypoglycaemia (dangerously low blood sugar) with sulfonylureas or insulin. Mitigation: more frequent glucose self-monitoring for the first month after starting or stopping.

  • Antihypertensives (blood-pressure drugs — angiotensin-converting enzyme inhibitors, angiotensin receptor blockers, beta-blockers): Monitor. Additive blood-pressure lowering reported in a hypertensive stress trial (Pattojoshi et al., 2026). Mitigation: home blood-pressure readings for four weeks after starting, with dizziness on standing noted.

  • Hepatotoxic drugs (acetaminophen, methotrexate, isoniazid, amiodarone): Caution. Overlapping liver toxicity raises the consequence of an injury event. Mitigation: baseline and 8–12 week liver panel; an alternative agent removes the overlap where one exists.

  • CYP3A4 and CYP2B6 substrates (statins — cholesterol-lowering drugs; direct oral anticoagulants — blood thinners; ciclosporin, bupropion, efavirenz): Caution, theoretical. Withanolides inhibit both enzymes in laboratory systems, potentially raising substrate levels. Mitigation: alertness to exaggerated drug effects; no human interaction study exists.

  • Over-the-counter sedating antihistamines (diphenhydramine, doxylamine) and alcohol: Caution. Additive drowsiness and impaired psychomotor performance. Mitigation: separation onto different evenings, and no alcohol within four hours of an ashwagandha dose.

  • Over-the-counter analgesics containing acetaminophen: Monitor. Additive hepatic burden, relevant at habitual rather than occasional use. Mitigation: acetaminophen below 2 g daily while supplementing, with liver enzymes in routine monitoring.

  • Sedative and calming supplements (valerian, kava, melatonin, magnesium glycinate, 5-HTP — 5-hydroxytryptophan, a serotonin building block — and L-Theanine): Caution; additive sedation, and for 5-HTP additive serotonin effect. These compound ashwagandha’s own action. Mitigation: staggered timing or a reduced dose of the co-agent.

  • Hepatotoxic botanicals (green tea extract, Garcinia cambogia, high-dose turmeric, red yeast rice, kava): Caution. Each carries an independent liver-injury signal; co-exposure confounds causality and raises risk. Mitigation: only one novel botanical at a time.

  • Other interventions — surgery and general anaesthesia: Caution. Additive sedation with anaesthetic agents and unknown effects on perioperative bleeding. Mitigation: discontinuation at least two weeks before an elective procedure.

Populations who should avoid Ashwagandha:

  • Solid-organ transplant recipients and anyone on immunosuppressive therapy, at any time post-transplant
  • People who are pregnant, attempting to conceive, or breastfeeding
  • People with any chronic liver disease — cirrhosis of any severity (Child-Pugh Class A through C, the standard liver-function grading), chronic hepatitis B or C, or fatty liver disease with scarring
  • People with a baseline alanine aminotransferase above twice the upper limit of normal, or any raised bilirubin
  • People with hyperthyroidism, Graves’ disease, or an autoimmune thyroid condition on active treatment
  • People with hormone-sensitive prostate cancer, given the documented testosterone rise in men
  • People with autoimmune disease under immune-suppressing treatment — rheumatoid arthritis, lupus, multiple sclerosis, inflammatory bowel disease
  • People with a prior episode of drug-induced or herb-induced liver injury from any agent
  • Anyone within two weeks of elective surgery under general anaesthesia

Risk Mitigation Strategies

  • Baseline liver panel before starting: Protocols cover alanine aminotransferase, aspartate aminotransferase, alkaline phosphatase and total bilirubin. This gives the comparator without which a later enzyme rise cannot be attributed, and screens out the pre-existing liver disease behind every fatal case.

  • Liver enzymes rechecked at 8–12 weeks: Reported injury latency is two to twelve weeks, so a single check inside that window catches most cases while they are still asymptomatic and fully reversible. Annual repetition follows with continued use.

  • Immediate discontinuation on jaundice, dark urine or new itching: These are the presenting features of ashwagandha cholestatic liver injury, not vague symptoms. Stopping at first appearance is what separates the self-limiting cases from the ones that progress.

  • Root-only extract: Predicted liver and reproductive toxicity is higher for leaf and aerial parts than for root, and several implicated products contained leaf material. Labels specifying root, rather than “whole plant” or “root and leaf”, identify these.

  • Two-week run-in at 300 mg daily: Half the standard dose for the first fortnight limits gastrointestinal upset and daytime sedation, the two most common adverse effects, and reveals individual sensitivity before full exposure.

  • Evening dosing with food: Food blunts nausea, cramping and reflux; evening timing moves the sedative peak away from driving and work hours and aligns it with the sleep benefit.

  • No concurrent novel botanical: Green tea extract, Garcinia cambogia, turmeric and red yeast rice each carry liver-injury signals. Single-agent introduction preserves the ability to attribute any adverse event to a cause.

  • Thyroid-stimulating hormone at 8 weeks on thyroid replacement: The small thyroxine-raising effect becomes clinically visible only in this group and can produce over-replacement symptoms at an unchanged prescription dose.

  • Third-party withanolide assay before purchase: Most commercial products fail pharmacopoeial content standards and many contain undisclosed leaf material. A lot-specific certificate of analysis confirming withanolide content and root-only sourcing prevents both wasted exposure and unexpected potency.

Therapeutic Protocol

  • Standard dose: 600 mg daily of a root extract standardized to at least 5% withanolides is the dose most trials used and where sleep, stress and strength effects were clearest. 300 mg daily is the common lower-tier protocol.

  • Extract selection: Practitioners generally specify a named standardized extract — KSM-66 (root only, roughly 5% withanolides) or Sensoril (root and leaf, roughly 10%) — because the trial evidence attaches to these, not to generic root powder.

  • Conventional versus integrative framing: Integrative practitioners position ashwagandha as a first-line adaptogen for stress-driven complaints; sleep-medicine and psychiatric practice treats it as an unvalidated option alongside cognitive behavioural therapy for insomnia. Neither framing is default.

  • Attribution of approaches: The 600 mg root-extract protocol traces to Indian trial groups around D.Y. Patil University School of Medicine and to Ixoreal Biomed’s KSM-66 programme; the low-dose, higher-withanolide approach traces to Natreon’s Sensoril and Arjuna’s Shoden.

  • Best time of day: Evening dosing is the common protocol, matching the sedative peak to bedtime and the cortisol-lowering effect to the natural evening trough. Split morning-and-evening dosing is used when daytime stress is the target.

  • Half-life and dosing frequency: Withanolide plasma half-life is only a few hours, which argues for split dosing, yet trials using 300 mg twice daily and 600 mg once daily report similar results because the effect accrues over weeks.

  • Single versus split dose: Split 300 mg twice daily is preferred where gastrointestinal upset appears or where daytime anxiety is the target; single 600 mg evening dosing is preferred where sleep is the target or adherence is the constraint.

  • Sex-based differences: Protocols are not dose-differentiated by sex, but expectations differ — men may see testosterone and dehydroepiandrosterone sulphate effects, women will not, and the perimenopausal indication uses the same 600 mg daily schedule.

  • Age-related considerations: At the older end of the target range, practitioners commonly start at 300 mg and stay there, because multiple concurrent medications, less liver reserve and fall risk from sedation all raise the cost of an adverse effect.

  • Baseline biomarkers influencing response: Higher baseline perceived-stress and anxiety scores, poorer baseline sleep, and lower baseline testosterone in men each predict a larger response; those with normal values at entry typically see little.

  • Pre-existing conditions influencing response: Subclinical hypothyroidism amplifies the thyroid response; insomnia amplifies the sleep response; active resistance training is a precondition for the strength response, which does not appear in untrained people.

  • Genetic polymorphisms influencing dose: No validated pharmacogenetic marker guides ashwagandha dosing. CYP3A4 and CYP2B6 variants plausibly alter withanolide clearance and co-medication interaction risk, but no trial has stratified on them, so dosing remains empirical.

Discontinuation & Cycling

  • Intended duration: Not established as lifelong. Almost every efficacy trial ran 8 to 12 weeks; the only long-term data is a 12-month safety cohort (Salve et al., 2026), so beyond a year has safety but not efficacy support.

  • Withdrawal effects: None documented. No trial has reported rebound anxiety, insomnia or hormonal overshoot on stopping, and no physical dependence syndrome has been described in the case literature.

  • Tapering protocol: No taper is described in the literature and none appears necessary given the absence of withdrawal effects. Abrupt discontinuation is the standard approach, and is mandatory if liver symptoms appear.

  • Cycling for efficacy: Whether effects diminish with continuous use is unknown. Because withanolides act on neurotransmitter receptors, tolerance cannot be ruled out; common practice is 8–12 weeks on followed by 2–4 weeks off.

  • Cycling for safety: Scheduled breaks also reset the liver-injury observation window and give an opportunity to recheck enzymes off-agent, which is the main practical argument for cycling in the absence of tolerance data.

Sourcing and Quality

  • Root-only extracts: Products specifying Withania somnifera root are the reference. Predicted liver and reproductive toxicity is higher for leaf and aerial material, and several products implicated in liver-injury cases contained leaf; “whole plant” and “root and leaf” labelling does not qualify.

  • Withanolide standardization: A stated withanolide percentage — commonly 5% for root-only extracts and 10% for root-and-leaf — marks a standardized product. Unstandardized “root powder” has no matching trial evidence and delivers unpredictable amounts of the active compounds.

  • Third-party testing: Independent testing found most products under-delivering withanolides and some exceeding heavy-metal limits. A lot-specific certificate of analysis covering withanolide assay plus lead, cadmium, arsenic and mercury documents the difference.

  • Named trial-grade extracts: KSM-66 (Ixoreal Biomed), Sensoril (Natreon) and Shoden (Arjuna Natural) are the extracts used in the published trials. Choosing one of these is the only way to inherit the trial evidence rather than assume it.

  • Certification marks: United States Pharmacopeia verification, NSF Certified for Sport and Informed Sport marks each add independent identity and contaminant testing; the sport marks additionally screen for undeclared substances, useful for tested athletes.

Practical Considerations

  • Time to effect: Sleep and anxiety changes typically emerge at two to four weeks and reach their measured size at eight weeks. Strength and cognitive effects were measured at eight weeks. Nothing meaningful happens acutely.

  • Common pitfall — unstandardized product: Buying by milligrams of powder rather than milligrams of withanolides is the most frequent error, and explains a large share of reported non-response given how many products under-deliver.

  • Common pitfall — stacking botanicals: Starting ashwagandha alongside other supplements removes the ability to attribute either benefit or an adverse event, and raises liver risk where the co-agents carry their own hepatotoxicity signals.

  • Common pitfall — indefinite use without monitoring: Because liver injury appears at two to twelve weeks and is silent at first, continuous use without a liver panel forfeits the one window in which the harm is fully reversible.

  • Regulatory status: Sold as a dietary supplement in the United States, so not reviewed by the Food and Drug Administration for safety or efficacy before sale. Denmark restricts it in food supplements; it is not prohibited under anti-doping rules.

  • Cost and accessibility: Neither expensive nor hard to obtain — trial-grade standardized extracts run roughly $15–30 monthly. Because it is self-paid and not reimbursed, no insurer or health system has a financial incentive favouring or opposing it against prescription anti-anxiety or sleep drugs.

Interaction with Foundational Habits

  • Sleep: Direct and potentiating. Ashwagandha shortens sleep-onset latency and improves sleep efficiency, with the largest effects in poor sleepers, and its sedative action on calming neurotransmitter receptors peaks two to four hours after dosing. Practical consequence: evening dosing, and expect compounding with any other sedating agent taken at night.

  • Nutrition: Indirect. No nutrient depletion is documented. Withanolides are fat-soluble, and the traditional preparation with milk or clarified butter plausibly aids absorption, though no human bioavailability study confirms this. Taking the dose with a fat-containing meal is the practical version, and it also blunts the early nausea and cramping.

  • Exercise: Direct and potentiating for resistance training specifically. Strength and muscle-size gains appear only in people already training, and creatine kinase — a marker of muscle damage — rises less after sessions. Aerobic capacity also improves. Practical consequence: it amplifies a training stimulus rather than substituting for one, and dose timing around sessions appears irrelevant.

  • Stress management: Direct and potentiating, acting on the same brain-to-adrenal pathway that meditation, breathwork and deliberate recovery modulate. Pooled data show cortisol falling substantially (Fornalik et al., 2026). Practical consequence: effects overlap rather than add indefinitely, and the largest benefits were recorded in people with high baseline stress scores, not in the already well-regulated.

Monitoring Protocol & Defining Success

Before starting, the literature’s central recommendation is a liver panel — alanine aminotransferase, aspartate aminotransferase, alkaline phosphatase and total bilirubin — because the reported liver injury cannot be attributed without a pre-exposure comparator, and because pre-existing liver disease was the factor separating self-limiting cases from fatal ones. A thyroid panel and, for men, morning total and free testosterone establish the two hormonal axes the herb measurably moves. Baseline perceived-stress and sleep-quality scores make the primary benefit measurable rather than impressionistic.

For ongoing monitoring, recheck liver enzymes at 8–12 weeks — the reported injury window — then annually with continued use. Recheck thyroid-stimulating hormone at 8 weeks in anyone on thyroid replacement or with borderline baseline values, and testosterone at 12 weeks in men tracking that outcome. The stress and sleep questionnaires are repeated at 4 and 8 weeks; absent change by then is the signal to stop.

Biomarker Optimal Functional Range Why Measure It? Context/Notes
Alanine aminotransferase 10–26 U/L (men), 8–22 U/L (women) Primary detector of the bile-flow and mixed liver injury reported with ashwagandha Conventional labs flag only above ~40–55 U/L; functional practitioners act far earlier. Non-fasting is acceptable. Best paired with aspartate aminotransferase and alkaline phosphatase
Aspartate aminotransferase 10–26 U/L Confirms a liver-cell injury pattern and separates liver from muscle sources Conventional range runs to ~40 U/L. Rises with intense exercise and with muscle damage, so heavy training within 48 hours of the draw confounds it. The informative reading is its ratio to alanine aminotransferase, not either value alone
Alkaline phosphatase 45–90 U/L The bile-flow marker; reported ashwagandha injury blocks bile flow, so this often moves first Conventional upper limit runs to ~120 U/L. A fasting sample is preferred. Best paired with gamma-glutamyl transferase, a bile-duct enzyme, which confirms a liver rather than bone source
Total bilirubin 0.3–1.0 mg/dL Jaundice is the presenting sign in the published case series; a rise marks a serious rather than incidental injury Fasting sample; rises transiently with fasting and with the harmless inherited variant Gilbert’s syndrome. Any new rise on ashwagandha is the reported trigger for immediate discontinuation
Thyroid-stimulating hormone 0.5–2.0 mIU/L Detects the small thyroid-stimulating effect, which is clinically visible only in those already on replacement or borderline Conventional range extends to 4.5 mIU/L. Drawn before 10 a.m. and before any thyroid medication dose; best paired with free thyroxine
Free thyroxine 1.0–1.5 ng/dL Pooled data show thyroxine specifically rising with ashwagandha, so this is the parameter that actually moves Conventional range is wider, roughly 0.8–1.8 ng/dL. Drawn with thyroid-stimulating hormone in the same sample. Free triiodothyronine adds value where symptoms of over-replacement appear
Total and free testosterone (men) Total 600–900 ng/dL; free 15–25 pg/mL The documented male-specific hormonal response and the marker of the vitality-related benefit Conventional total-testosterone range starts far lower, around 264–916 ng/dL. Drawn between 7 and 10 a.m. when levels peak; two separate mornings give a reliable baseline. Not informative in women, where no effect is seen
Morning serum cortisol 10–15 µg/dL at 8 a.m. The proposed mechanism marker; falls substantially in pooled trial data Conventional morning range is much wider, roughly 6–23 µg/dL. Drawn at a fixed time; a single reading is noisy. A four-point salivary curve is the functional-medicine alternative and reflects the daily rhythm better
Fasting glucose 75–86 mg/dL Ashwagandha modestly lowers fasting glucose, which matters when combined with insulin or sulfonylureas Conventional range extends to 99 mg/dL before anything is flagged. Requires a 12-hour fast. Best paired with fasting insulin or with glycated haemoglobin, a three-month average of blood sugar, for context
Complete blood count Within conventional reference limits, no individual downward trend Screens for the blood-cell changes assessed in the randomized safety trials (Verma et al., 2021); none were found, making any deviation notable No fasting required. The functional target here is stability against the person’s own baseline rather than a distinct optimal band

Qualitative markers worth tracking alongside the labs:

  • Sleep-onset latency and number of night awakenings, recorded nightly rather than recalled weekly
  • Perceived stress and irritability, ideally on a repeated validated questionnaire rather than free recall
  • Morning energy and daytime drowsiness — the latter distinguishes benefit from over-sedation
  • Cognitive clarity, sustained attention and working memory during demanding tasks
  • Training performance: session quality, next-day soreness and recovery between hard sessions
  • Libido and sexual function in men, where the hormonal response is measurable
  • Gastrointestinal comfort, particularly in the first two weeks when upset is most likely
  • Any yellowing of skin or eyes, dark urine, unexplained itching or right-upper-abdominal discomfort — the trigger for immediate discontinuation

Emerging Research

  • Non-restorative sleep trial: NCT07085819 is recruiting 160 adults for a four-arm dose-ranging test of an ashwagandha root-plus-leaf aqueous extract against placebo, with restorative sleep quality as the primary endpoint — targeting unrefreshing sleep rather than insomnia.

  • Diurnal cortisol in women: NCT07647081 will follow 30 women over eight weeks, measuring change in the daily salivary cortisol curve and heart-rate variability. It directly tests whether the proposed mechanism holds in a female-only population where hormonal effects have differed.

  • Skin and hair endpoints: NCT07215689 is recruiting 50 healthy adults on 300 mg twice daily, with transepidermal water loss (a measure of skin-barrier integrity) as its primary endpoint — a new outcome domain that could broaden the benefit profile or return a null.

  • Withaferin A in graft-versus-host disease: NCT07708987 is a phase 2 study of standardized withaferin A in 34 patients whose graft-versus-host disease (transplanted immune cells attacking the recipient’s tissues) has not responded to steroids — the first serious test of the isolated compound as a drug.

  • Menopausal symptom replication: NCT07210229 will randomize 90 women across three arms using the Menopause Rating Scale and menopause-specific quality of life, providing the independent replication the current single-trial perimenopause evidence lacks.

  • Hepatotoxicity surveillance could weaken the case: Population exposure estimates put roughly 15.6 million United States adults on one of six liver-implicated botanicals in any 30-day window (Likhitsup et al., 2024). Prospective registry work is the route to an actual incidence rate.

  • Root-versus-leaf toxicity modelling: Computational structure-activity work argues root is materially safer than leaf and aerial parts (Ronen et al., 2025). Experimental confirmation would either narrow the safety concern to specific products or extend it to all of them.

  • Endocrine reassessment: A 2026 evaluation re-examines the hormonal-activity claims that drove European regulatory restriction (Vollmer & Brendler, 2026). Its conclusions will shape whether restrictions spread beyond Denmark or are relaxed.

Conclusion

Ashwagandha is a root extract from Indian traditional medicine whose measurable action is to soften the body’s stress response. For readers who treat stress load, sleep and physical capacity as levers on long-term health, the clearest returns are in those three domains: calmer self-reported stress and anxiety, faster sleep onset and better sleep efficiency in poor sleepers, and added strength and aerobic capacity when it accompanies actual training. Effects on memory and attention are consistent though less certain, and men see a rise in testosterone that has not yet been tied to any outcome that matters clinically. Nothing here has been shown to extend human healthspan; the longevity case rests on worms and cell cultures.

Set against this is a genuine and serious liver signal. Independent hospital groups on three continents have described the same pattern of bile-flow liver injury appearing within weeks of starting, mostly reversible but fatal in people whose livers were already damaged. A smaller thyroid effect matters mainly for those already on thyroid medication, and transplant recipients face outright danger.

The evidence base is also compromised at its foundation. Most positive trials were paid for by the companies selling the extracts, they are short, and independent testing shows many commercial products contain far less active material than their labels claim. The safety literature, by contrast, comes from hospital groups with nothing to sell.

Top - Benefits - Risks - Protocol