Reviewed records available

Ashwagandha

Withania somnifera

The human evidence is more interesting than a simple yes or no. Five small sleep trials produced a favorable pooled signal. A separate stress review found lower cortisol but no statistically significant improvement in perceived stress.

Evidence snapshot reviewed October 16, 2026

Check the botanical identity in the Plant Atlas →

Botanical identity

Which plant this profile means

This profile is tied to Withania somnifera while preserving plant-part, extraction, withanolide, dose, and duration differences.

Scientific name
Withania somnifera
WCVP taxon ID
2465599
Atlas record
Open this name in the Plant Atlas

Preparation boundaries

Same plant, different exposure

These forms stay separate in the evidence bank. A result for one cannot be copied onto the others.

mixed standardized extracts

root, leaf, or root-and-leaf · Taken by mouth

Trials use different plant parts, extraction methods, withanolide specifications, doses, and durations; the pooled evidence cannot establish retail-product equivalence.

Research-bot status

Some evidence cards are public. The rest of the queue isn’t.

The reviewed cards below cleared a separate publication check. The remaining candidate extractions stay private until their source, wording, and preparation match are reviewed.

Sources in review
2
Possible claims found
3
Still awaiting review
0
Blocked source
0

Research connections

What this ingredient has been studied around

Sleep

A reviewed, preparation-specific evidence card is published for this source.

Stress

A reviewed, preparation-specific evidence card is published for this source.

Reviewed evidence

2 source-linked records

Evidence synthesis Promising, with limits 5 trials with 400 adults

Ashwagandha extracts and sleep: a small pooled signal

Five small trials produced a pooled improvement in overall sleep, with a larger signal in the insomnia subgroup. That is encouraging, not definitive. The products varied, heterogeneity was substantial, and the review said serious-adverse-effect and long-term safety data were limited.

What was studied
Oral ashwagandha extracts from different plant parts and specifications; retail-product equivalence was not established
Setting
Studied alone
Why it ranks here
The randomized evidence points in a favorable direction, but it rests on five trials, 400 participants, mixed extracts, and limited long-term safety data.
Limits and cautions
  • Only five trials and 400 adults were included.
  • The pooled result had substantial heterogeneity.
  • Root, leaf, and mixed-part extracts with different specifications cannot be assumed equivalent.
  • No serious adverse effects were reported in the included trials, but the review said the data were too limited to establish long-term safety.
  • The subgroup findings do not establish an ideal dose or duration.

Ashwagandha is not appropriate for everyone. Pregnancy, breastfeeding, thyroid or autoimmune disease, upcoming surgery, liver concerns, and medicines that affect sedation, blood pressure, blood sugar, thyroid function, or immunity deserve clinician or pharmacist review.

Evidence synthesis Mixed or uncertain 488 participants represented across cortisol and perceived-stress analyses

Ashwagandha and stress: cortisol moved, perceived stress did not

The pooled trials found lower cortisol, but they did not find a statistically significant improvement in perceived stress. That mismatch matters: a biomarker change is not the same as feeling better or functioning better. The studies also had moderate risk of bias and did not settle long-term safety, dosing, or product equivalence.

What was studied
Mixed oral ashwagandha extracts using different plant parts, withanolide specifications, doses, and durations
Setting
Studied alone
Why it ranks here
The biological marker moved while the patient-reported outcome did not, so a broad claim that ashwagandha reduces stress would outrun the data.
Limits and cautions
  • Cortisol is a biomarker and cannot substitute for a patient-reported or functional outcome.
  • The perceived-stress result was not statistically significant.
  • Risk of bias was generally moderate, with some allocation-concealment problems.
  • The pooled products used different plant parts, extract specifications, doses, and durations.
  • Long-term safety and effectiveness across diverse populations remain unresolved.

This record does not establish ashwagandha as a treatment for anxiety, depression, trauma, or another mental-health condition. Persistent or severe symptoms deserve qualified care.

Profile sources

Read the source material