Research dossier
Clinical research on Ashwagandha
9 trials reviewed across 4 indications.
Strongest evidence
Stress and cortisol reduction
Mechanism
Withanolides modulate HPA-axis tone — dampening cortisol output and shifting the cortisol-to-DHEA balance. The herb is best understood as an HPA-axis brake, not a stimulant or sedative.
Among adaptogens, ashwagandha has the deepest RCT bench: multiple double-blind trials show 15–30% drops in perceived stress and meaningful cortisol reductions at 240–600 mg/day of standardized extract over 8–10 weeks. Effects are largest in adults already running high stress.
Effect size is largest in chronically stressed adults. Healthy non-stressed adults see smaller changes — there is no extra cortisol to push down.
Trials cited
Chandrasekhar — KSM-66 ashwagandha for stress
positive · RCT
Chandrasekhar et al., 2012, Indian Journal of Psychological Medicinen=6464 adults with chronic stress took 600 mg/day of full-spectrum root extract (commonly cited as KSM-66) for 60 days. Perceived Stress Scale fell sharply versus placebo (p<0.0001) and serum cortisol dropped (p=0.0006). The most-cited ashwagandha stress trial.
Modest sample size and a population specifically selected for chronic stress — effects in non-stressed adults will be smaller.
Salve — ashwagandha dose-finding for stress
positive · RCT
Salve, Pate, Debnath, Langade, 2019, Cureusn=60Industry-funded60 healthy adults assigned to 250 mg/day, 600 mg/day, or placebo for 8 weeks. Both doses cut PSS scores (p<0.05 at 250 mg, p<0.001 at 600 mg) and lowered cortisol versus placebo, with 600 mg/day stronger across endpoints. Dose-response is real but plateaus quickly.
Single-site, manufacturer-supported study. Replication outside industry-funded settings is still limited.
Lopresti — Shoden ashwagandha for stress and HPA-axis tone
positive · RCT
Lopresti et al., 2019, Medicine (Baltimore)n=60Industry-funded60 stressed adults on 240 mg/day Shoden saw significant Hamilton Anxiety reductions (p=0.040), greater morning cortisol drops, and a rise in DHEA-S versus placebo over 60 days. Authors attribute the effect to HPA-axis modulation.
Industry-supported (Arjuna Natural). Branded extract dose is not interchangeable with generic root powder.
Pratte — systematic review of ashwagandha for anxiety
positive · Systematic review
Pratte et al., 2014, Journal of Alternative and Complementary MedicineSystematic review screened 62 abstracts and pooled 5 RCTs. All five reported greater anxiety or stress score improvement on ashwagandha than placebo. The signal is consistent across small trials — but the underlying studies are small and methodologically uneven.
Authors flag heterogeneous methods and possible bias in component trials. Treats the herb as anxiolytic-promising, not anxiolytic-proven.
Sleep onset and quality
Mechanism
Sleep benefits likely flow downstream of stress and HPA-axis effects rather than direct GABAergic action. Lower evening cortisol and reduced sympathetic tone make falling asleep easier.
Two trials from the Langade group show 5–8% gains in sleep efficiency and meaningful drops in sleep onset latency in adults with poor sleep. Effects in healthy good sleepers are smaller. Not a sedative — closer to a slow-acting anxiolytic that improves sleep secondarily.
Useful in stress-driven insomnia. Not appropriate as a same-night sleep aid the way melatonin or doxepin are.
Langade — ashwagandha for insomnia and anxiety
positive · RCT
Langade, Kanchi, Salve, Debnath, Ambegaokar, 2019, Cureusn=58Industry-funded60 adults with poor sleep took 600 mg/day for 10 weeks. Sleep onset latency dropped (29.0 vs 33.9 min, p=0.019), sleep efficiency rose (75.6 → 83.5 vs 75.1 → 79.7), and Hamilton Anxiety scores improved versus placebo.
Self-reported sleep diaries plus PSQI rather than polysomnography. Effect on sleep latency is small in absolute terms.
Langade — actigraphy-confirmed sleep effects
positive · RCT
Langade, Thakare, Kanchi, Kelgane, 2021, Journal of Ethnopharmacologyn=80Industry-fundedTwo-arm trial measuring sleep with wrist actigraphy. In the insomnia arm, ashwagandha significantly improved sleep onset latency (p<0.0001), sleep efficiency (p<0.0001), total sleep time (p<0.002), and WASO (p=0.040) versus placebo at 8 weeks.
Actigraphy is closer to objective than diaries but is not full polysomnography. Healthy-sleeper arm showed smaller effects — most of the benefit lives in people who already sleep poorly.
Salve — ashwagandha dose-finding for stress
positive · RCT
Salve, Pate, Debnath, Langade, 2019, Cureusn=60Industry-funded60 healthy adults assigned to 250 mg/day, 600 mg/day, or placebo for 8 weeks. Both doses cut PSS scores (p<0.05 at 250 mg, p<0.001 at 600 mg) and lowered cortisol versus placebo, with 600 mg/day stronger across endpoints. Dose-response is real but plateaus quickly.
Single-site, manufacturer-supported study. Replication outside industry-funded settings is still limited.
Strength and muscle growth in trained men
Mechanism
Plausible candidates include reduced cortisol-driven catabolism, modest testosterone elevation, and improved exercise recovery. The mechanism is not fully resolved.
One well-designed RCT in resistance-training novices found ashwagandha roughly doubled strength gains and modestly raised testosterone over 8 weeks. The effect is real in beginners; it has not been cleanly replicated in trained lifters.
Strongest effect in untrained or new-to-training men. Trained athletes should expect smaller gains and weaker effect sizes.
Wankhede — ashwagandha for strength and muscle in trained men
positive · RCT
Wankhede et al., 2015, Journal of the International Society of Sports Nutritionn=57Industry-funded57 men starting resistance training took 600 mg/day ashwagandha or placebo for 8 weeks. Bench-press 1-RM rose 46.0 kg vs 26.4 kg (p=0.001), leg extension 14.5 vs 9.8 kg (p=0.04), and serum testosterone climbed 96 ng/dL vs 18 ng/dL (p=0.004).
Beginners gain quickly with any program, which inflates between-group differences. Strength gains in trained lifters are likely smaller. Industry-supported.
Thyroid and androgen effects
Mechanism
Ashwagandha appears to push thyroid hormone output up — useful in subclinical hypothyroidism, dangerous in hyperthyroidism. Effects on testosterone and DHEA are modest and inconsistent across trials.
In subclinical hypothyroid adults, 600 mg/day normalized TSH and raised free T3 and T4 over 8 weeks. Several stress trials report parallel rises in testosterone and DHEA-S. The same hormonal activity that helps a sluggish thyroid is exactly the wrong intervention for an overactive one.
Avoid in hyperthyroidism, Graves', or any thyroid-stimulating-antibody picture. Discuss with your prescriber if you take levothyroxine — your dose may need re-titration.
Sharma — ashwagandha in subclinical hypothyroidism
positive · RCT
Sharma, Basu, Singh, 2018, Journal of Alternative and Complementary Medicinen=5050 adults with subclinical hypothyroidism on 600 mg/day saw TSH normalize (p<0.001), free T3 rise (p<0.001), and free T4 rise (p=0.0096) versus placebo over 8 weeks. The same mechanism is the warning label: ashwagandha pushes thyroid hormone up, which is a risk in hyperthyroid or thyroid-medicated patients.
Single small site; subclinical population only — hyperthyroid or Graves' patients should not extrapolate. The thyroid effect is precisely why people on levothyroxine need to talk to their prescriber before starting.
Lopresti — Shoden ashwagandha for stress and HPA-axis tone
positive · RCT
Lopresti et al., 2019, Medicine (Baltimore)n=60Industry-funded60 stressed adults on 240 mg/day Shoden saw significant Hamilton Anxiety reductions (p=0.040), greater morning cortisol drops, and a rise in DHEA-S versus placebo over 60 days. Authors attribute the effect to HPA-axis modulation.
Industry-supported (Arjuna Natural). Branded extract dose is not interchangeable with generic root powder.
Wankhede — ashwagandha for strength and muscle in trained men
positive · RCT
Wankhede et al., 2015, Journal of the International Society of Sports Nutritionn=57Industry-funded57 men starting resistance training took 600 mg/day ashwagandha or placebo for 8 weeks. Bench-press 1-RM rose 46.0 kg vs 26.4 kg (p=0.001), leg extension 14.5 vs 9.8 kg (p=0.04), and serum testosterone climbed 96 ng/dL vs 18 ng/dL (p=0.004).
Beginners gain quickly with any program, which inflates between-group differences. Strength gains in trained lifters are likely smaller. Industry-supported.
Honest-evidence ledger — 1 trial that didn’t move the needle
Surfacing failed trials alongside the positive evidence. Leaving them out would be marketing, not science.
Björnsson — ashwagandha-induced liver injury case series
negative · Observational
Björnsson et al., 2020, Liver Internationaln=5Five patients developed cholestatic or mixed-pattern liver injury after starting ashwagandha. Latency averaged 2–12 weeks. All recovered after stopping the herb. The signal is rare in absolute terms but mechanistically real and reproducible across continents.
Case-series evidence cannot establish incidence rate. Genetic susceptibility, undisclosed contaminants, and product-quality variability all contribute. Treat any new jaundice or right-upper-quadrant pain on ashwagandha as a stop-and-test signal.
4 forms of Ashwagandha compared
KSM-66 (Ixoreal Biomed)
KSM-66
Standardized to ≥5% withanolides; full-spectrum root only
Best forStress, sleep, strength — the most-studied branded extractRoot-only, no leaf, milk-extracted with no chemical solvents. The Chandrasekhar 2012 trial is the most-cited result and used a comparable 600 mg/day full-spectrum root extract.
stress300–600 mgsleep300–600 mgmuscle600–600 mgSensoril (Natreon)
Sensoril
Standardized to ≥10% withanolides; root + leaf blend
Best forStress, anxiety — second most-studied branded extractHigher withanolide concentration than KSM-66 due to leaf inclusion, which means lower mg doses (125–500 mg/day) can hit therapeutic withanolide loads. Effective per-mg, but the leaf component carries more withaferin-A — a more cytotoxic withanolide.
stress125–500 mgShoden (Arjuna Natural)
Shoden
Concentrated to 35% withanolide glycosides — highest in class
Best forStress, HPA-axis — works at low mg dosesThe Lopresti 2019 trial used 240 mg/day Shoden and saw significant anxiety and cortisol effects. High potency means dose-creep risks are real if formulators don't adjust the mg load downward.
stress60–240 mgGeneric ashwagandha root extract
Variable — withanolide content rarely disclosed
Best forGeneral use; quality depends entirely on the supplierIf the label does not specify a withanolide percentage or branded extract, you do not know what you are taking. Quality varies widely between contract manufacturers.
Side effects and drug interactions
Side effects
GI upset and nausea
Common
Most common side effect — usually mild, often resolves with food or splitting the dose. Reported in 5–10% of trial participants.
Drowsiness
Uncommon
Some users feel sedated, especially at higher doses or when taken in the morning. Move dosing to evening if this happens.
Thyroid hormone elevation
Uncommon
Free T3 and T4 rise across multiple trials. Welcome in subclinical hypothyroidism, dangerous in hyperthyroidism, and a dose-stability problem on levothyroxine.
Liver injury
Rare
Cholestatic or mixed-pattern hepatitis with jaundice has been reported across at least two independent case series (Björnsson 2020, Lubarska 2023). Recovery on cessation. Latency 2–12 weeks. Rare in absolute terms but mechanistically real.
Reproductive and pregnancy effects
Severe
Traditionally classified as abortifacient at high doses in animal data. No human pregnancy safety data. Avoid in pregnancy.
Drug interactions
Additive effect
sedativesbenzodiazepinesz-drugs (zolpidem, eszopiclone)alcoholAdditive sedation. Ashwagandha's calming effect compounds with CNS depressants.
Watch for excessive drowsiness. Do not combine with sleep medications without clinician input.
Additive effect
levothyroxineliothyroninethyroid extractAshwagandha raises endogenous thyroid hormone. Stacked with replacement therapy, this can push patients into iatrogenic hyperthyroidism.
Talk to your prescriber before starting. TSH should be re-checked at 4–8 weeks if you proceed.
Other
immunosuppressantsTNF-alpha inhibitorsDMARDsAshwagandha is mildly immunostimulatory. In autoimmune disease (Hashimoto's, lupus, MS, RA), this can theoretically flare disease activity.
Avoid in active autoimmune disease unless specifically cleared by your specialist.
Additive effect
antihypertensivesantidiabetic medicationsModest blood-pressure-lowering and glucose-lowering effects in some trials. Additive risk with prescription therapy.
Monitor home blood pressure and glucose for the first 2–4 weeks.
Other critical caveats
- Avoid in pregnancy. There is no human safety data and traditional use plus animal data suggest abortifacient potential.
- Do not start ashwagandha alongside levothyroxine or in hyperthyroidism without prescriber input — the herb pushes thyroid hormone up.
- Stop immediately and seek medical evaluation for any new jaundice, dark urine, persistent right-upper-quadrant pain, or unexplained fatigue. Liver injury cases are rare but documented.
- Avoid in active autoimmune disease (lupus, RA, Hashimoto's, MS). The mild immunostimulatory effect can theoretically worsen disease activity.
- Most published RCTs are industry-funded. The signal across stress and cortisol is consistent enough to take seriously, but expect smaller effects than headline numbers when you actually take it.
Frequently asked
Which ashwagandha form should I buy?
If a product names KSM-66, Sensoril, or Shoden, you know what is in the capsule. KSM-66 has the deepest stress-and-strength evidence base at 300–600 mg/day. Sensoril works at 125–500 mg. Shoden works at 60–240 mg. A label that just says 'ashwagandha extract' without a withanolide percentage is a black box — pass.How much should I take?
For stress and sleep: 300–600 mg/day of KSM-66 or equivalent withanolide load from another branded extract. For strength alongside resistance training: 600 mg/day. Take with food if it upsets your stomach. Effects build over 4–8 weeks — this is not a same-day intervention.Will ashwagandha mess with my thyroid?
Yes — that is one of its real effects. In Sharma 2018, 600 mg/day normalized TSH and raised free T3 and T4 in subclinical hypothyroid adults. That is helpful if you are slightly underactive, harmful if you are overactive, and a dose-stability problem if you take levothyroxine. Talk to your prescriber before starting.Is ashwagandha safe long-term?
Most trials run 8–12 weeks. There is no long-term RCT safety data past 6 months. Liver injury cases are rare but documented across multiple countries. A reasonable approach is to use it for defined stress periods, take regular breaks, and stop if you notice fatigue, jaundice, or right-upper-quadrant discomfort.Does ashwagandha boost testosterone?
Modestly, in some trials. Wankhede 2015 saw a ~96 ng/dL rise in untrained men starting resistance training. Lopresti 2019 saw smaller androgen shifts in stressed adults. The effect is real but small — ashwagandha is not a TRT replacement and the effect size in trained men is unclear.Can I take ashwagandha during pregnancy?
No. There is no human pregnancy safety data and traditional Ayurvedic use classifies it as abortifacient at high doses. Animal studies support the concern. Avoid throughout pregnancy and during attempts to conceive if the timing is sensitive.
References
Last reviewed2026-05-07