Supplement Signals

Ashwagandha

Root · Stress & sleep

A root from traditional Indian medicine, studied mostly as extracts. In most reviews and trials it lowered stress and anxiety more than placebo, mostly in stressed adults, and cortisol in most of the studies that measured it. It improved sleep and added strength with training; for testosterone in men the evidence is preliminary.

Thiago’s take: Situational · 600 mg every evening

Ashwagandha (Withania somnifera) is a plant of the nightshade family with a thousand-year history in traditional Indian medicine. Its main compounds are withanolides, steroid-like lactones: the trials’ extracts stated 1.5 to 8% of them where one review could tell, while most extracts sold are not standardised to them.

My take

  1. Core
  2. Situational
  3. Trying
  4. Researching
  5. Not taking

A-tier. Not something you need, but the one I’d add to wind down and sleep better.

Dose
600 mga day
Timing
Eveningwith dinner
Since
20233 years
Brand I use
Santa Cruz Paleo

Where it fits my day

  1. Morning with breakfast

  2. Training 30 minutes before

    • Electrolytes
    • HMB

  3. Evening with dinner

  4. Bed 30 minutes before

    • Sleep blend

If you’re starting

  1. What to buy Read the label for three things: the part of the plant (root, or root and leaf, both used in trials), whether it is an extract or a powder, and its withanolide content, which the trials’ extracts mostly state. Mine says root, 600 mg, and gives no withanolide content.
  2. How to take it I take one 600 mg capsule after dinner, as I wind down for bed. The newest stress trials gave 30 to 600 mg of extract a day, in one dose or two, for one to two months.
  3. What to expect Not an overnight change. In stressed adults, stress scores fell more than on placebo within two to eight weeks in the trials, and one review that pooled perceived stress alone found no clear difference.

What the studies show

4 Strong1 Preliminary

Each segment is one studyBetterNo changeWorse

Strong: at least one pooled review or three randomised trials among the counted studies, and at least three in four of them better. Preliminary: everything else. Better is green in a Strong finding and ochre in a Preliminary one.

Better means a statistically significant change in the claimed direction at a normal dose. Each study counts once: a pooled review speaks for the trials inside it.

Stress and anxiety, in stressed or anxious adults

1,002 people, Akhgarjand 2022 · lower stress and anxiety, the stress effect best at 300 to 600 mg a day

7 of 8 studies better

In adults, most of them selected for high stress or anxiety, ashwagandha extracts lowered stress and anxiety scores more than placebo in a review of 12 trials and in four newer trials at 30 to 600 mg a day, though a review that pooled the Perceived Stress Scale alone found no significant drop. In people with anxiety disorders, insomnia and other mental disorders it lowered anxiety scores too, the reviews rating their trials low in certainty and very varied.

Strong

Stress

What these 8 studies found. Pooled reviews: 3 better, 1 no change. Trials: 4 better.

  1. BetterPubMed

    In adults aged 25 to 48, anxiety (SMD −1.55) and stress (SMD −1.75) lower than placebo, with high heterogeneity and low-certainty evidence; the stress effect was best at 300 to 600 mg a day.

    Akhgarjand 2022 Phytotherapy Research

    • Meta-analysis
    • 12 trials · 1,002 people
  2. BetterPubMed

    In people with mental disorders, mostly anxiety disorders, anxiety scores lower than the comparators (SMD −1.13 without the outlier trial at 12,000 mg a day, 9 of its 14 trials), confirmed in a Bayesian model, and perceived stress lower (SMD −0.95, 3 trials); it rates the evidence low in certainty.

    Marchi 2025 BJPsych Open

    • Meta-analysis
    • 14 trials · 713 people
    • 10 to 12,000 mg a day, median 600 mg
  3. No changePubMed

    In adults, perceived stress (PSS) no lower than placebo (SMD −0.36, p = 0.40), at 250 mg a day or more for two weeks or longer; cortisol, pooled separately, fell by 1.16 µg/dL.

    Albalawi 2025 Nutrition and Health

    • Meta-analysis
    • 488 adults
    • 250 mg a day or more
  4. BetterPubMed

    In people with anxiety or insomnia, anxiety scores (HAM-⁠A) 6.0 points lower than placebo, with very high heterogeneity between its five trials (I² 98%).

    Fatima 2024 Human Psychopharmacology

    • Meta-analysis
    • 5 trials · 254 people
  5. BetterPubMed

    In its ashwagandha and placebo arms (122 of 186 highly stressed adults), 175 mg of a full-spectrum ashwagandha root extract twice a day for 60 days lowered perceived stress (PSS) by 4.7 points more than placebo, and DASS-⁠21 anxiety and stress scores more; DASS-⁠21 depression scores and mental alertness did not differ.

    McKinney 2026 Trials

    • Randomised trial
    • 186 highly stressed adults aged 18 to 65
    • 175 mg ashwagandha root extract twice a day · 60 days
  6. BetterPubMed

    In its ashwagandha root extract and placebo arms (94 of 141 adults with moderate stress and anxiety), 300 mg twice a day for 8 weeks lowered perceived stress (PSS) by 8.2 points against 1.3 and anxiety (HAM-⁠A) by 8.2 against 1.6, more than placebo; serum cortisol, its main outcome, did not differ.

    West 2026 Journal of Medicine and Life

    • Randomised trial
    • 141 adults aged 18 to 65 with moderate stress and anxiety
    • 300 mg ashwagandha root extract twice a day · 8 weeks
  7. BetterPubMed

    In 135 healthy adults with moderate perceived stress, 150 or 300 mg of a sustained-release ashwagandha root extract a day for 60 days left perceived stress (PSS) lower than placebo, 13.3 and 13.2 against 18.5, and sleep quality scores better.

    Thanawala 2026 Medicine

    • Randomised trial
    • 135 healthy adults aged 20 to 55 with moderate perceived stress
    • 150 or 300 mg sustained-release ashwagandha root extract a day · 60 days
  8. BetterPubMed

    In 60 adults with self-reported stress, 30 mg of an ashwagandha extract (Ashwa.30) a day for 28 days lowered self-reported stress more than placebo in the full analysis set (p = 0.046), not significantly per protocol, and daily stress ratings and the stress rated during a cold-pressor test on day 28 (p = 0.044) too; its other secondary and physiological outcomes, cortisol during a stress test aside, did not differ.

    Lopresti 2026 Advances in Therapy

    • Randomised trial
    • 60 adults aged 18 to 65 with self-reported stress
    • 30 mg ashwagandha extract (Ashwa.30) a day · 28 days

Found with a PubMed search: ashwagandha or Withania in the title with stress or anxiety in the title or abstract, or ashwagandha or Withania in the title or abstract with stress, anxiety or adaptogen in the title, reviews and trials, human studies. 78 results on 5 Oct 2026.

Cortisol

1,706 people, Fornalik 2026 · lower cortisol, perhaps more with more withanolides

4 of 5 studies better

In adults under stress, a review found cortisol in blood fell in all seven of its studies that measured it, though mostly against people’s own starting levels, and a newer review of stress trials pooled a drop of 1.16 µg/dL against placebo; in adults in placebo-controlled trials of hormone levels, a 2026 review found it lower too. In two trials of stressed adults, the largest found serum cortisol no lower than with placebo after 8 weeks at 600 mg a day, and one newer than every review found a smaller cortisol rise during a stress test at 30 mg a day.

Strong

Stress

What these 5 studies found. Pooled reviews: 3 better. Trials: 1 better, 1 no change.

  1. BetterPubMed

    In adults in placebo-controlled trials of hormone levels, cortisol lower than placebo (SMD −1.18), with subgroups suggesting a dose response by withanolide content.

    Fornalik 2026 Planta Medica

    • Meta-analysis
    • 23 trials · 1,706 people
  2. BetterPubMed

    In healthy adults under stress, cortisol fell from each group’s starting level in all seven of its studies that measured it in blood, by 11 to 33%, one of them a controlled study without placebo; the two that measured it in saliva disagreed. It finds a cortisol-lowering, stress-reducing effect over 30 to 112 days, and notes the time of day was rarely accounted for.

    Della Porta 2023 Nutrients

    • Systematic review
    • 673 healthy adults under stress
    • 60 mg to 5 g a day
  3. BetterPubMed

    In adults in trials that measured cortisol and perceived stress, cortisol 1.16 µg/dL lower than placebo (95% CI −1.64 to −0.69) across its seven studies of cortisol, at 250 mg a day or more for two weeks or longer; perceived stress, pooled separately, did not differ.

    Albalawi 2025 Nutrition and Health

    • Meta-analysis
    • 488 adults
    • 250 mg a day or more
  4. No changePubMed

    In its ashwagandha root extract and placebo arms (94 of 141 adults with moderate stress and anxiety), 300 mg twice a day for 8 weeks lowered serum cortisol, its main outcome, by 2.6 µg/dL against 1.1, not significantly (p = 0.13).

    West 2026 Journal of Medicine and Life

    • Randomised trial
    • 141 adults aged 18 to 65 with moderate stress and anxiety
    • 300 mg ashwagandha root extract twice a day · 8 weeks
  5. BetterPubMed

    In 60 adults with self-reported stress, 30 mg of an ashwagandha extract (Ashwa.30) a day for 28 days kept salivary cortisol steady during a cold-pressor stress test on day 28, while it rose on placebo (p = 0.047).

    Lopresti 2026 Advances in Therapy

    • Randomised trial
    • 60 adults aged 18 to 65 with self-reported stress
    • 30 mg ashwagandha extract (Ashwa.30) a day · 28 days

Found with a PubMed search: ashwagandha or Withania in the title with cortisol in the title or abstract, or ashwagandha or Withania in the title or abstract with cortisol, the stress (HPA) axis, the adrenals or stress hormones in the title, reviews and trials, human studies. 20 results on 5 Oct 2026.

Sleep

400 people, Cheah 2021 · a small gain in sleep, larger in insomnia and at 600 mg a day or more

8 of 8 studies better

In adults with or without insomnia, a review of five trials found ashwagandha extracts improved sleep a little more than placebo, more so in insomnia, and two reviews of people with anxiety, insomnia and other mental disorders found sleep scores better too, one rating the evidence low in certainty. Five newer trials, in adults with sleep disturbance, stressed adults, older adults and postmenopausal women, at 150 to 600 mg a day, found sleep or daytime sleepiness scores better than on placebo, though some single measures, such as the time spent asleep in one, did not differ.

Strong

Sleep

What these 8 studies found. Pooled reviews: 3 better. Trials: 5 better.

  1. BetterPubMed

    In people with mental disorders, sleep quality better than placebo (SMD −1.35) in 4 of its 14 trials, two of them in insomnia and one giving ashwagandha with piperine; it rates the evidence low in certainty.

    Marchi 2025 BJPsych Open

    • Meta-analysis
    • 14 trials · 713 people
    • 10 to 12,000 mg a day, median 600 mg
  2. BetterPubMed

    In adults with or without insomnia, overall sleep better than placebo (SMD −0.59), a small effect, larger in insomnia (2 trials) and at 600 mg a day or more, with root extracts and, in one trial, a root-and-leaf extract; sleep quality ratings, the time to fall asleep, total sleep time, time awake after falling asleep and sleep efficiency improved, while PSQI scores (2 trials) and time in bed did not differ.

    Cheah 2021 PLOS ONE

    • Meta-analysis
    • 5 trials · 400 people
    • 120 to 600 mg a day
  3. BetterPubMed

    In people with anxiety or insomnia, the time to fall asleep, total sleep time, sleep efficiency and PSQI scores better than placebo; time awake after falling asleep and time in bed did not differ.

    Fatima 2024 Human Psychopharmacology

    • Meta-analysis
    • 5 trials · 254 people
  4. BetterPubMed

    In its ashwagandha and placebo arms (100 of 200 adults with sleep disturbance), 300 mg of a root extract (KSM-⁠66) twice a day for 8 weeks shortened the time to fall asleep and lengthened total sleep time on a wrist monitor, and improved time awake after falling asleep, sleep efficiency, rated sleep quality and PSQI scores, more than placebo; mental alertness on rising did not differ, and time in bed differed only at week 4.

    Movva 2026 Clocks & Sleep

    • Randomised trial
    • 200 adults aged 18 to 50 with sleep disturbance
    • 300 mg ashwagandha root extract (KSM-⁠66) twice a day · 8 weeks
  5. BetterPubMed

    In its ashwagandha and placebo arms (122 of 186 highly stressed adults, people with sleep disorders excluded), 175 mg of a full-spectrum ashwagandha root extract twice a day improved PSQI scores by 1.6 points and restorative sleep (RSQ-⁠W) by 8.9 points more than placebo at day 60, not at day 30; the PSQI’s sleep duration and sleep disturbance parts did not differ.

    McKinney 2026 Trials

    • Randomised trial
    • 186 highly stressed adults aged 18 to 65
    • 175 mg ashwagandha root extract twice a day · 60 days
  6. BetterPubMed

    In 135 healthy adults with moderate perceived stress, 150 or 300 mg of a sustained-release ashwagandha root extract a day left PSQI scores lower than placebo from day 7, at 5.5 and 4.5 against 7.6 by day 60.

    Thanawala 2026 Medicine

    • Randomised trial
    • 135 healthy adults aged 20 to 55 with moderate perceived stress
    • 150 or 300 mg sustained-release ashwagandha root extract a day · 60 days
  7. BetterPubMed

    In 66 postmenopausal women aged 45 to 65, 600 mg of an ashwagandha root extract a day for 12 weeks lowered PSQI scores by 5.2 points against 0.8 on placebo (P < 0.001).

    Obhrai 2026 Menopause

    • Randomised trial
    • 66 postmenopausal women aged 45 to 65
    • 600 mg ashwagandha root extract a day · 12 weeks
  8. BetterPubMed

    In 50 healthy adults aged 60 to 85, 300 mg of an ashwagandha root extract twice a day for 8 weeks showed “higher improvements” in daytime sleepiness (Epworth Sleepiness Scale) after week 8 (p < 0.001), and quality of life, its main outcome, improved more than with placebo.

    Honnutagi 2026 Phytotherapy Research

    • Randomised trial
    • 50 healthy adults aged 60 to 85
    • 300 mg ashwagandha root extract twice a day · 8 weeks

Found with a PubMed search: ashwagandha or Withania in the title with sleep or insomnia in the title or abstract, or ashwagandha or Withania in the title or abstract with sleep or insomnia in the title, reviews and trials, human studies. 39 results on 5 Oct 2026.

Testosterone in men

1,706 men and women in all, Fornalik 2026 · testosterone 57 ng/dL higher in men, with no significance given for men alone

1 of 3 studies better

In men in placebo-controlled trials of hormone levels, a 2026 review found testosterone about 57 ng/dL higher with ashwagandha than with placebo, though it gives no significance for men alone. In the largest trial, of 100 otherwise healthy men selected for sexual dissatisfaction, 600 mg a day for 8 weeks raised testosterone by 120 ng/dL more than placebo, while a newer trial of recreationally active adults found the men’s rise no greater than on placebo, their levels staying within the normal range.

Preliminary

Hormones

What these 3 studies found. Pooled reviews: 1 no change. Trials: 1 better, 1 no change.

  1. No changePubMed

    In adults in placebo-controlled trials of hormone levels, testosterone in men 57 ng/dL higher than placebo (MD 57.43 ng/dL) and not in women, the difference between the sexes significant; the abstract gives no significance for men alone, so it is graded no change.

    Fornalik 2026 Planta Medica

    • Meta-analysis
    • 23 trials · 1,706 people
  2. BetterPubMed

    In 100 otherwise healthy men aged 30 to 50 selected for sexual dissatisfaction, 300 mg of an ashwagandha root extract twice a day for 8 weeks raised serum testosterone by 70 ng/dL while it fell by 50 on placebo, a difference of 120 ng/dL (p = 0.012); dihydrotestosterone did not differ.

    Mutha 2025 Journal of Ayurveda and Integrative Medicine

    • Randomised trial
    • 100 otherwise healthy men aged 30 to 50 with sexual dissatisfaction
    • 300 mg ashwagandha root extract twice a day · 8 weeks
  3. No changePubMed

    In 84 recreationally active adults aged 25 to 50, 300 mg of a sustained-release ashwagandha root extract a day for 90 days raised the men’s total and free testosterone from their starting levels, total by 121 ng/dL, but no more than placebo (p > 0.05), their levels staying within the normal range in both groups.

    Thanawala 2026 Frontiers in Nutrition

    • Randomised trial
    • 84 recreationally active adults aged 25 to 50
    • 300 mg sustained-release ashwagandha root extract a day · 90 days

Found with a PubMed search: ashwagandha or Withania in the title with testosterone or androgens in the title or abstract, or ashwagandha or Withania in the title or abstract with testosterone, androgens or hormones in the title, reviews and trials, human studies. 20 results on 5 Oct 2026.

Strength with training

161 people, Lee 2026 · bench-press and lower-body 1-⁠RM up more than with placebo, not significantly under its stricter sensitivity analysis

3 of 3 studies better

In adults in training, two 2026 reviews found ashwagandha extracts, root or root-and-leaf, raised strength more than placebo with the same training, though the one confined to structured resistance-training trials found the gain lost significance under a stricter analysis and rates the evidence low in certainty. A newer trial in 84 recreationally active adults new to strength training found bench-press strength rose 8 kg more than on placebo over 90 days at 300 mg a day of a sustained-release root extract.

Strong

Performance

What these 3 studies found. Pooled reviews: 2 better. Trials: 1 better.

  1. BetterPubMed

    In its four strength trials (223 of its 1,249 adults), each adding ashwagandha to the same training in both arms, muscle strength rose more than with placebo (SMD 0.58), significantly in physically active people (2 trials, SMD 1.03) and not in untrained or special populations (2 trials); it rates the evidence low in certainty.

    Zhu 2026 Frontiers in Pharmacology

    • Meta-analysis
    • 20 trials · 1,249 people
  2. BetterPubMed

    In healthy adults in structured resistance training, 500 to 600 mg a day of a root extract (KSM-⁠66, 2 trials) or a root-and-leaf extract (Sensoril, 1) raised bench-press 1-⁠RM (SMD 0.60) and lower-body 1-⁠RM (SMD 0.52) more than placebo with the same training in its main model, and chest girth (2 trials); body fat did not differ. Under its Hartung-Knapp sensitivity analysis both strength estimates lost significance, and it rates them low in certainty.

    Lee 2026 Nutrients

    • Meta-analysis
    • 3 trials · 161 people
    • 500 to 600 mg a day
  3. BetterPubMed

    In 84 recreationally active adults aged 25 to 50 new to strength training, on the same resistance and aerobic programme five days a week, 300 mg of a sustained-release ashwagandha root extract a day for 90 days raised bench-press 1-⁠RM, estimated from repetitions to exhaustion, by 8.0 kg more than placebo (p < 0.001), its main outcome, and VO2max more too; muscle-damage markers (CK, LDH) did not differ. The newest review (Lee 2026) could not read its bench-press values as absolute loads and left it unpooled.

    Thanawala 2026 Frontiers in Nutrition

    • Randomised trial
    • 84 recreationally active adults aged 25 to 50
    • 300 mg sustained-release ashwagandha root extract a day · 90 days

Found with a PubMed search: ashwagandha or Withania in the title with strength, resistance training, muscle, exercise, athletes or one-repetition maximum in the title or abstract, or ashwagandha or Withania in the title or abstract with strength, muscle, exercise, athletes, performance, training or physical in the title, reviews and trials, human studies. 27 results on 5 Oct 2026.

Dose

Thiago takes 600 mg of root a day, and his label states no extract and no withanolide content. The trials’ doses are of extracts standardised to their withanolides, from 1.5% to over 5% total withanolides in the ones that state it, most at 300 to 600 mg a day: one review found the stress effect best in that range and another the sleep effect larger at 600 mg or more, while one trial gave 30 mg of a low-dose extract.

Daily300 to 600 mg, most trials’ extract doses

Thiago, 600 mg

Sources: Akhgarjand 2022 · Cheah 2021 · Lee 2026 · West 2026 · Movva 2026 · Smith 2023 · Thanawala 2026 · Lopresti 2026

Safety

What people notice, an uncommon but serious liver injury, the thyroid, the regulators, and when to ask a doctor first.

  • CommonSleepiness and stomach upset

    The side effects reported most are mild and passing: sleepiness, stomach discomfort and diarrhoea. In trials in people with mental disorders they were no more common than with the comparators, and in 23 trials of 2,317 healthy adults on root extracts they were infrequent and seen on the comparators too.

    Sources: Della Porta 2023 · Marchi 2025 · Coope 2026

  • OccasionalLiver injury

    Uncommon but serious: a 2026 review found 25 people with liver injury in 13 reports, some on products mixing it with other ingredients, mostly cholestatic or mixed (bile flow impaired, with jaundice and itching) after weeks of use, and in two case series testing found ashwagandha and no contaminant. Most recovered over weeks to months after stopping; one needed a liver transplant, and three died, all of them people who already had cirrhosis, and the review advises people with advanced liver disease to avoid it.

    Sources: McIntyre 2026 · Björnsson 2020 · Philips 2023

  • OccasionalThyroid hormone

    In placebo-controlled trials a 2026 review found T4, a thyroid hormone, a little higher on ashwagandha (by 0.61 µg/dL), with TSH and T3 unchanged, and in 50 adults with a mildly underactive thyroid 600 mg of root extract a day improved TSH, T3 and T4 toward normal more than placebo. Three case reports describe an overactive thyroid (thyrotoxicosis) in people taking it, which eased after they stopped.

    Sources: Fornalik 2026 · Sharma 2018 · van der Hooft 2005 · Kamal 2022 · Hayashi 2024

  • UnprovenHarm to reproduction

    Denmark’s food authority banned it, citing effects on sex hormones and reproduction, and EU regulators have recommended a procedure that could lead to a ban in food supplements. A 2026 review finds the evidence that it harms reproduction inconclusive, the human studies it reviews showing no adverse effect on hormonal balance or reproductive health.

    Sources: Patwardhan 2024 · Brendler 2026 · Tallon 2026

  • Ask a doctor firstIf you are

    Pregnant. No placebo-controlled trial has been run in pregnancy: the one trial, of 70 women in their second trimester, was open-label against standard care and reported no adverse events in 12 weeks. A 2026 review calls the evidence that it harms pregnancy inconclusive, its reputation for causing miscarriage resting on distorted citations, and another advises restricting it in pregnant women.

    Sources: Ajgaonkar 2026 · Tallon 2026 · Li 2026

  • Ask a doctor firstIf you take

    Thyroid medicine, or sedatives and sleeping pills. No study in people has tested ashwagandha alongside them, and a 2025 review calls for studies of its use with other medicines.

    Source: Marchi 2025

Forms

  • Root powder

    The dried root, ground, rather than an extract of it, and Thiago’s form.

    This page

  • KSM-66

    A root-only extract, made without alcohol or chemical solvents, with over 5% withanolides.

    Movva 2026 · Cheah 2021

  • Sensoril

    A standardised water extract of the root and leaf.

    Lee 2026

  • Shoden

    An extract of the root and leaf standardised to 35% withanolide glycosides, given at 120 mg a day in a sleep trial.

    Lopresti 2019 · Cheah 2021

  • Witholytin

    A water-and-alcohol extract of the root, standardised to 1.5% withanolides.

    Smith 2023

  • Sustained-release root extract

    A root extract with at least 4% withanolides, in a capsule made to release them slowly and taken once a day.

    Thanawala 2026 · Thanawala 2026

  • Ashwa.30

    A low-dose extract, given at 30 mg a day in a 2026 stress trial, where most trials have used 250 to 600 mg.

    Lopresti 2026

How it works

Ashwagandha’s withanolides are thought to act on the stress axis, which releases cortisol, though how it lowers cortisol or helps sleep is not yet known.

  1. ~140compounds in the plant

    Ashwagandha, a nightshade with a thousand-year history in traditional Indian medicine, holds withanolides among them, steroids that also occur as glycosides. A 2023 review attributes most of its effects to two, withaferin A and withanolide D.

  2. HPAthe stress axis

    Under stress the hypothalamus signals the pituitary, which signals the adrenal glands to release cortisol. Withanolides are thought to act on this axis, and on inflammation and oxidation.

  3. 11–33%less cortisol in blood

    In seven studies of stressed adults that measured it in blood, cortisol fell from where it began. How is not yet clear: perhaps withaferin A binds the receptor cortisol acts on, or its sedative effect lowers it.

  4. TEGfor sleep, in mice

    In mice an extract rich in withanolides failed to bring on sleep, while water-soluble parts, triethylene glycol (TEG) among them, did. How it helps sleep in people is not known.

Sources: Della Porta 2023 · Lee 2026 · Cheah 2021