Ashwagandha (Withania somnifera (L.) Dunal) is promoted for stress, anxiety, sleep, cognition, hormones, thyroid function, and exercise performance. The evidence is uneven: most trials are small, short, and tied to a specific preparation, so findings from one extract or powder cannot automatically be transferred to another product.
The Ayurvedic Pharmacopoeia of India identifies Aśvagandhā as the dried mature root. It lists tikta and kaṣāya rasa, laghu guna, uṣṇa vīrya, madhura vipāka, and the actions rasāyana, vātakaphāpaha, balya, and vājīkaraṇa. Its powder dose is 3–6 g. This is not equivalent to the milligram dose of a concentrated extract, and classical properties should not be converted directly into claims for modern diagnosed diseases.
Moderate Evidence: Stress and Cortisol Reduction
Stress is the best-supported modern use, but “strong evidence” overstates certainty. A 2022 meta-analysis in Phytotherapy Research included 12 randomized-trial papers with 1,002 participants. Ashwagandha reduced stress (standardized mean difference −1.75; 95% CI −2.29 to −1.22) and anxiety (−1.55; 95% CI −2.37 to −0.74), but heterogeneity was high and certainty was rated low.
A 2021 review covered seven trials involving 491 adults treated for six to eight weeks. Studies used root extracts, root-and-leaf extracts, an unspecified extract, or root granules; extract doses ranged from 240 to 1,250 mg/day, while the granules equaled 6 g/day of root powder. Most reported improved perceived stress, and several reported lower cortisol, but the preparation differences prevent a universal dose recommendation.
In the Salve trial, 60 stressed adults received aqueous root extract at 125 mg twice daily, 300 mg twice daily, or placebo for eight weeks. Both active regimens improved stress measures and reduced cortisol. Confidence level: Moderate because samples were small, follow-up was short, and products and methods varied.
Limited-Moderate Evidence: Anxiety Reduction
Anxiety scores often improve in stress trials, but many participants had self-reported stress or mild symptoms rather than a rigorously diagnosed anxiety disorder. NCCIH states that some preparations may help stress and insomnia while evidence for anxiety remains unclear.
An international psychiatric taskforce provisionally discussed 300–600 mg/day of standardized root extract for generalized anxiety disorder while emphasizing that stronger evidence is needed. Ashwagandha should not replace psychotherapy, psychiatric assessment, or prescribed treatment. Confidence level: Limited to moderate.
Moderate Evidence: Sleep Quality
A 2021 meta-analysis included five randomized trials with 400 participants and found a small but significant overall sleep benefit (standardized mean difference −0.59; 95% CI −0.75 to −0.42). Effects were more prominent in adults with insomnia, at doses of at least 600 mg/day, and when treatment lasted at least eight weeks.
Four trials used KSM-66 root extract and one used Shoden root-and-leaf extract. Outcomes included sleep quantity, quality, latency, efficiency, and alertness on rising. Certain extracts may improve sleep, but they are not proven substitutes for evaluating chronic insomnia, sleep apnea, depression, medication effects, or thyroid disease. Confidence level: Moderate.
Limited-Moderate Evidence: Muscle Strength and Recovery
The 2015 Wankhede trial enrolled 57 healthy men aged 18–50 with little resistance-training experience. Participants received 300 mg of root extract twice daily or placebo during eight weeks of training. Mean bench-press strength rose by 46.0 kg with ashwagandha and 26.4 kg with placebo; leg-extension strength and selected muscle-size measures also improved more in the active group.
The trial additionally reported differences in testosterone, body-fat percentage, and serum creatine kinase. A 2026 meta-analysis pooled four strength studies involving 223 participants and found a favorable effect, but certainty for muscle strength was low and heterogeneity was moderate; pooled body weight and body-fat effects were not significant.
Confidence level: Limited to moderate; the findings do not justify broad claims for all athletes or preparations.
Limited Evidence: Testosterone and Male Fertility
In a 2019 crossover trial, overweight men aged 40–70 with mild fatigue received standardized Shoden extract for eight weeks. Compared with placebo, DHEA-S increased by 18% more and testosterone by 14.7% more.
The study found no significant between-group differences in fatigue, vigor, sexual well-being, psychological well-being, cortisol, or estradiol. A 2026 meta-analysis of seven testosterone studies involving 488 participants found a small pooled increase, but populations and preparations differed. NCCIH describes testosterone and sperm-quality evidence as limited.
Ashwagandha is not an established treatment for hypogonadism, erectile dysfunction, or infertility, and results from stressed, overweight, training, or subfertile populations should not be generalized to every healthy man. Confidence level: Limited and population-specific.
Preliminary Evidence and Caution: Thyroid Function
A double-blind trial in 50 adults with subclinical hypothyroidism compared 300 mg of root extract twice daily with placebo for eight weeks. TSH decreased and T3 and T4 increased in the active group. One small trial does not establish treatment for overt hypothyroidism, Hashimoto disease, hyperthyroidism, thyroid nodules, or patients stabilized on thyroid medicine.
NIH sources also describe thyroid-hormone changes and case reports of thyrotoxicosis that improved after ashwagandha was stopped. People with thyroid disease, abnormal tests, or thyroid medication should not use it without clinician supervision. Confidence level: Preliminary for benefit, with a meaningful safety signal.
Limited-Moderate Evidence: Cognition and Memory
A 2017 pilot trial randomized 50 adults with mild cognitive impairment to 300 mg of root extract twice daily or placebo for eight weeks. The active group improved on tests of immediate and general memory, executive function, sustained attention, and information-processing speed. This does not establish prevention or treatment of dementia or another neurological disease.
A 2026 meta-analysis included 20 studies with 1,249 participants across cognitive and physical outcomes. Pooled results favored ashwagandha for memory and attention or processing speed; certainty was moderate for these outcomes and low for executive function. Global cognition came from one study and was rated very low certainty. Confidence level: Limited to moderate for selected test outcomes.
Insufficient Evidence: Blood Sugar and Inflammation
A 2020 diabetes review found only five clinical studies alongside much more laboratory and animal research and concluded that human evidence was not robust enough for therapeutic use. Small studies have also measured inflammatory or oxidative-stress biomarkers, but biomarker changes do not prove treatment of inflammatory disease. NCCIH considers evidence insufficient for diabetes and most other promoted conditions.
Root Powder and Proprietary Extracts Are Not Interchangeable
“Ashwagandha” is not one uniform intervention. The pharmacopoeial drug is dried mature root; KSM-66 is a root extract; and Sensoril and Shoden are root-and-leaf extracts used at different doses and standardizations. Evidence belongs first to the exact preparation and regimen studied.
| Preparation | Verified distinction | Main limitation |
|---|---|---|
| API root powder | Dried mature root; dose 3–6 g | Not directly convertible to extract milligrams |
| KSM-66 | Root extract used in multiple trials | Results apply to comparable regimens |
| Sensoril | Root-and-leaf extract used in stress studies | Not identical to the official root drug |
| Shoden | Concentrated root-and-leaf extract | Lower mass does not prove superiority |
| Unspecified extract | Plant part or standardization may be unclear | Difficult to match reliably to research |
A stated withanolide percentage is not a universal potency scale. Plant part, extraction, marker definition, dose, accompanying ingredients, duration, product quality, and patient population all affect whether research applies to a commercial supplement.
Safety Summary
Ashwagandha may be reasonably tolerated for short-term use, generally up to about three months, but long-term safety is unknown. Reported effects include drowsiness, stomach upset, diarrhea, and vomiting.
Rare clinically apparent liver injury has been reported. LiverTox describes typical onset two to twelve weeks after starting ashwagandha, usually with cholestatic or mixed injury, jaundice, and itching. Most cases improve after stopping, but liver failure, transplantation, and deaths have occurred, particularly with pre-existing liver disease or cirrhosis. Avoid use in cirrhosis or advanced chronic liver disease and avoid rechallenge after suspected injury.
Important cautions and possible interactions include:
- Pregnancy and breastfeeding: NCCIH advises avoidance.
- Thyroid disorders or thyroid medicine: Hormone changes may interfere with treatment.
- Sedatives and anticonvulsants: Additive nervous-system effects are possible.
- Diabetes and blood-pressure medicines: Additive glucose- or pressure-lowering effects are possible.
- Immunosuppressants or autoimmune disorders: Potential immune effects require caution.
- Surgery: NCCIH does not recommend use when surgery is approaching.
- Hormone-sensitive prostate cancer: Avoid use because testosterone may increase.
Stop use and seek medical assessment for jaundice, dark urine, severe itching, persistent vomiting, unusual fatigue, upper-abdominal discomfort, palpitations, tremor, or marked heat intolerance.
Dosage by Indication: Evidence Reference
No universal therapeutic dose is established. These are selected study regimens and the pharmacopoeial powder dose, not interchangeable prescriptions.
| Outcome studied | Example regimen | Limitation |
|---|---|---|
| Perceived stress | 125 or 300 mg aqueous root extract twice daily for 8 weeks | One product-specific trial |
| Stress/anxiety studies | Various extracts, 240–1,250 mg/day for 6–8 weeks | Preparations and scales varied |
| Sleep | Product-specific extracts; stronger subgroup signal at ≥600 mg/day and ≥8 weeks | Not a universal dose |
| Resistance training | 300 mg root extract twice daily for 8 weeks | Key trial involved inexperienced men |
| Cognition | 300 mg root extract twice daily for 8 weeks | Small pilot; other trials used other products |
| Subclinical hypothyroidism | 300 mg root extract twice daily for 8 weeks | Single trial plus thyroid risk |
| Traditional root powder | API: 3–6 g | Not an extract conversion |
The Bottom Line for Practitioners
The most defensible conclusion is that certain ashwagandha preparations may reduce perceived stress and modestly improve sleep during short-term use. Evidence for diagnosed anxiety disorders is less secure. Muscle strength, testosterone, cognition, male fertility, and thyroid effects remain limited, product-specific, or population-specific; evidence for treating diabetes and inflammatory disease is insufficient.
Responsible practice requires identifying the plant part and preparation, matching claims and doses to the tested product, and not assuming that short trials establish indefinite safety. Screen for pregnancy, breastfeeding, liver or thyroid disease, autoimmune disorders, surgery, hormone-sensitive prostate cancer, and interacting medicines. Consult a qualified Ayurvedic practitioner and an appropriate healthcare provider before medical use or use with prescription treatment.
This article is educational and does not diagnose, treat, or replace individualized care. Persistent anxiety, insomnia, fatigue, fertility concerns, cognitive symptoms, or abnormal thyroid, liver, glucose, or hormone results require professional assessment.
References
- Ayurvedic Pharmacopoeia of India
- Does Ashwagandha supplementation have a beneficial effect on the management of anxiety and stress? A systematic review and meta-analysis of randomized controlled trials (2022), PubMed
- NIH Office of Dietary Supplements
- Adaptogenic and Anxiolytic Effects of Ashwagandha Root Extract in Healthy Adults: A Double-blind, Randomized, Placebo-controlled Clinical Study (2019), PubMed
- Effect of Ashwagandha (Withania somnifera) extract on sleep: A systematic review and meta-analysis (2021), PubMed
- Examining the effect of Withania somnifera supplementation on muscle strength and recovery: a randomized controlled trial (2015), PubMed
- Cancer Stem Cells and Neovascularization (2021), PubMed
- Frontiersin (frontiersin.org)
- A Randomized, Double-Blind, Placebo-Controlled, Crossover Study Examining the Hormonal and Vitality Effects of Ashwagandha ( Withania somnifera) in Aging, Overweight Males (2019), PubMed
- NCCIH
- Efficacy and Safety of Ashwagandha Root Extract in Subclinical Hypothyroid Patients: A Double-Blind, Randomized Placebo-Controlled Trial (2018), PubMed
- Efficacy and Safety of Ashwagandha (Withania somnifera (L.) Dunal) Root Extract in Improving Memory and Cognitive Functions (2017), PubMed
- Withania somnifera (Indian ginseng) in diabetes mellitus: A systematic review and meta-analysis of scientific evidence from experimental research to clinical application (2020), PubMed
- NCBI
The evidence-stratified format is exactly what practitioners need when evaluating whether to recommend Ashwagandha for a specific indication. I’ve been using this herb in my practice since 2015 and the evidence base has grown significantly since then.
The honest assessment of limitations is what makes this different from every sponsored Ashwagandha article. Acknowledging that the testosterone data is modest and mostly in men with existing deficiency is more useful than claiming it raises testosterone for everyone.
The KSM-66 versus Sensoril versus raw extract distinction is something everyone taking Ashwagandha should understand. The standardization differences mean you’re not comparing equivalent products when you compare studies. This article handles that correctly.
The anxiety and sleep evidence is the strongest category and my clinical experience agrees. The cortisol studies and the GABAergic mechanism hypothesis together explain why the clinical effects are reliable in this indication even without complete mechanistic proof.
I liked the practical side of Ashwagandha. I would like to know how long to try it before judging results.
Is there any data on long-term use beyond 18 months? Most trials are 8-12 weeks. I’ve been taking Ashwagandha for three years and I want to know if there are any safety concerns that emerge with extended use.
The safest part of the Ashwagandha advice is keeping it simple. A few more examples would still help.
The safest part of the Ashwagandha advice is keeping it simple. The practical details matter more than people think.
The article makes a clear point that ashwagandha root powder dose is not interchangeable with concentrated extract milligrams.
The immune system evidence for Ashwagandha is categorized as preliminary here. I’d push back slightly: the NK cell activation data is more consistent than ‘preliminary’ suggests. The quality of the trials is the issue rather than the consistency of the findings.
I wonder how the heterogeneity in those meta-analyses affects confidence in the stress reduction results.
The safest part of the Ashwagandha advice is keeping it simple. Would be useful to see a short checklist next.
It would be helpful to see a comparison of KSM-66 versus Shoden extracts in sleep trials.
The breakdown of Ayurvedic properties like tikta rasa and madhura vipāka adds nice context for practitioners.
I find the caution about liver injury important, especially for anyone with pre existing conditions.
The mention that ashwagandha should not replace psychotherapy for anxiety feels responsible.
Did the authors note any specific timeline for when liver toxicity tends to appear?
I appreciate that the piece emphasizes matching product preparation to the studied regimen.
The limited evidence for testosterone changes makes me cautious about recommending it for fertility support.
It seems the best supported use remains short term stress reduction, but even that comes with low certainty.
The table contrasting root powder, KSM-66, Sensoril, and Shoden clarifies why dosing is so tricky.
I would like to see more long term safety data before considering ashwagandha for chronic use.
The evidence base for the fatigue protocol in section 4 seems thin. Are there actual RCTs or just observational studies?
Does thyroid interaction note interact with metformin? Would appreciate a separate post on herb-drug combos.
At what point should someone stop self-administering 300mg vs 600mg trial and see a vaidya?
At what point should someone stop self-administering KSM-66 vs Sensoril comparison and see a vaidya?
Tried the the cortisol reduction data protocol for 3 weeks now, sleep is noticeably better.
For someone with Pitta imbalance, does 300mg vs 600mg trial need to be modified?
this actually worked for me! tried KSM-66 vs Sensoril comparison for 2 weeks, energy is better
finally tried 300mg vs 600mg trial after reading this. no miracle cure but genuinely helpful
You mention 300mg vs 600mg trial works for Vata types, but what about mixed Vata-Pitta constitutions?
Where are the actual studies on the fatigue protocol in section 4? A few citations would strengthen this considerably.
not convinced by the KSM-66 vs Sensoril comparison argument. would need to see better evidence
interesting but i disagree with the the cortisol reduction data approach. my experience was the opposite
My grandmother used to do exactly what you described for the cortisol reduction data. Never had a name for it until now.
How long before you start seeing results with the cortisol reduction data? My practitioner said 90 days.
Tried the thyroid interaction note protocol for 3 weeks now, sleep is noticeably better.
Is there an age limit for thyroid interaction note? Asking for my 68-year-old father.
Sharing this with my mother, she has been struggling with exactly what the the cortisol reduction data section describes.
Been following this blog for a while and the KSM-66 vs Sensoril comparison post is one of the most useful ones.
the scientific evidence behind the fatigue protocol in section 4 seems to lag well behind the traditional claims.
Trying this next month.
What’s the difference between the KSM-66 vs Sensoril comparison approach for acute vs chronic conditions?
For someone with Pitta imbalance, does KSM-66 vs Sensoril comparison need to be modified?
not convinced by the thyroid interaction note argument. would need to see better evidence