Ashwagandha with antidepressants or anti-anxiety medication:
what has been tested, what has been reported and what you never do
There is exactly one randomised study that asked this question properly: it gave ashwagandha to people who were already taking an antidepressant. There were 40 of them and they were followed for six weeks. And there is a published case in which the combination went wrong, in a 22-year-old woman who had been taking escitalopram for two years. This article puts both on the table, plus the thing nobody says in Romanian: “antidepressants” and “anti-anxiety medication” are two different questions, with two different answers.
The short answer:
“antidepressants” and “anti-anxiety medication” are not the same question
Antidepressants and anti-anxiety medicines both end up in the same box in the bathroom, but they do not do the same thing and they do not touch the same systems. A selective serotonin reuptake inhibitor is taken daily, for months or years, and works on serotonin. A benzodiazepine is usually taken as needed, in the short term, and works on the GABA-A receptor. Ashwagandha has a different answer for each of them, and the two answers are nothing alike.
For the first half of the question there is a randomised study. For the second there is none, but there is a plausible mechanism and two case reports. This means that the half the Romanian internet writes about most, antidepressants, is in fact the half with the most data, while the half it says nothing about, anti-anxiety medication, is the one where an overlap is more likely.
| The two halves | Antidepressants | Benzodiazepine anti-anxiety medicines |
|---|---|---|
| Examples | sertraline, escitalopram, fluoxetine, venlafaxine, duloxetine | alprazolam, lorazepam, diazepam, bromazepam |
| How they are taken | daily, for months or years | usually as needed, in the short term |
| System affected | serotonin and noradrenaline | the GABA-A receptor |
| What has been tested with ashwagandha | one randomised study, 40 people | no study |
| What has been reported | one case of serotonin syndrome | one case of sedation, confusion and amnesia, with no other sedative |
If an article answers both halves with the same sentence, it has read neither the study nor the case report.
The only study that asked the question properly:
40 people already taking an SSRI
In 2021, an Iranian team published a randomised, double blind, placebo controlled study of 40 patients with generalised anxiety disorder diagnosed according to DSM-IV-TR criteria. The detail that makes the study relevant is easy to miss in the abstract: all the patients were already being treated with selective serotonin reuptake inhibitors. The extract was not tested instead of the medication, but on top of it.
Twenty-two received root extract, 1 g a day, one capsule. Eighteen received placebo. Six weeks, with Hamilton Anxiety Rating Scale assessments at the start, in week two and in week six. The score fell by 14 points in the extract group and by 8 points in the placebo group, and the difference was reported as statistically significant.
A study that measures whether something helps does not automatically measure whether it is safe. They are two questions, with two different kinds of study.
What the study does not say is at least as important. It does not say which SSRI the participants were taking or at what dose. It does not report adverse events in the abstract. It does not follow anyone after the six weeks. And, most importantly, 40 people over six weeks cannot detect a rare event. If a complication occurs in one in several thousand people, a study of this size will never see it. Its absence from the report is not proof that it does not exist.
The rest of the literature:
14 studies, 713 people, and an eight-week follow-up
In 2025 a systematic review with meta-analysis appeared on exactly the question “what does ashwagandha do in people with mental disorders”, not in stressed healthy volunteers. It included 14 randomised studies, 360 people treated with extract and 353 controls. The predominant diagnostic category was anxiety disorders. Thirteen of the studies gave the extract orally, at a median dose of 600 mg a day, and the median follow-up was 8 weeks.
The result, in short: the effect on anxiety symptoms held up under an additional Bayesian analysis, while the effect on depression did not. On tolerability, the authors found no significant difference between the extract and the comparators, and digestive complaints were the most common reactions, described as mild and transient. The authors list their own limitations: few studies, substantial heterogeneity between them, outlier effects.
| Type of evidence | What it answers | What it does not answer |
|---|---|---|
| The randomised study on SSRIs, 2021 | it helps on top of an antidepressant, over six weeks, in 40 people | safety, rare events, what happens afterwards |
| The 2025 meta-analysis | what happens in people with mental disorders, in general | what happens in people on medication, separately |
| Case reports | what can happen | how often it happens |
| Regulatory dossiers | what the authorities do with the product | what happens in your body |
None of the four lines answers the question in the title on its own. Together, they sketch something honest: a plant with a real signal on anxiety, a complete gap on interactions and a few isolated reports that show what can go wrong.
The mechanism the authors point to:
it is not serotonin, it is GABA
This is where the Romanian internet consistently gets it wrong. The explanation in circulation is that ashwagandha “acts on serotonin receptors”, so it clashes with the antidepressant. The authors of the only randomised study on this topic write something else, in the very introduction to their paper: it is a plant with a GABAergic property, and that is where its anxiolytic effect comes from. The same explanation appears in the 2025 case report on prolonged sleep and confusion, where the authors talk about GABAergic modulation by withanolides.
If the mechanism is GABA, the practical consequence is the reverse of what is written everywhere. The GABA-A receptor is exactly where benzodiazepines work. In other words, the theoretical overlap is not with the antidepressant, it is with the anti-anxiety medicine. And the US professional fact sheet lists sedatives among the medicines the plant may interact with, alongside antidiabetic medicines, antihypertensives and immunosuppressants. Antidepressants do not appear on that list at all.
This does not mean the serotonergic mechanism is impossible. It only means it is the hypothesis from a single case, not a demonstrated property of the plant. Both things can be true at the same time, and a serious article does not choose between them, it states both.
What has been tested: exactly your situation, once. Forty people with generalised anxiety disorder, all being treated with selective serotonin reuptake inhibitors, 22 on 1 g a day of extract and 18 on placebo, for six weeks. The anxiety score fell by 14 points in the extract group and by 8 in the placebo group.
What has not been tested: safety. A six-week study of 40 people is not designed to catch a rare event, and serotonin syndrome is a rare event. The second thing not tested: what happens after six weeks.
What to do: tell the doctor who prescribed your antidepressant, beforehand, not afterwards. And do not change anything about its dose.
What has been tested: nothing, directly. There is no study of people taking a benzodiazepine and ashwagandha at the same time.
What is known all the same: the authors of the study above describe the plant as having a GABAergic property, and benzodiazepines act on the GABA-A receptor. The US professional fact sheet lists sedatives among the medicines the plant may interact with. Drowsiness is a common effect of the plant on its own.
What to do: if you take the anti-anxiety medicine rarely, the practical question is what happens on the evening you take both. Do not drive or drink alcohol that evening, and tell your doctor.
What has been tested: nothing, again. This is where the gap is biggest, because it is also the situation in which an overlap in effect is most likely.
What is known: there is a case published in 2025 of prolonged sleep, abnormal sleep behaviours, acute confusion and retrograde amnesia in a 40-year-old man, on ashwagandha alone, with no other sedative. Everything resolved completely after stopping. The authors attribute it to GABAergic modulation by withanolides and to individual differences in metabolism.
What to do: do not add anything on top of a daily sedative regimen without your doctor. This is not a generic precautionary rule, it is the direct conclusion of that case.
What has been tested: enough for a discussion, too little for a promise. Fourteen randomised studies, 713 participants, median dose 600 mg a day, median follow-up 8 weeks.
What has not been tested: what happens beyond three months. And what happens if you start an antidepressant later, on top of the plant.
What to do: if your symptoms are bad enough that you are thinking about medication, the right order is a consultation first. A supplement is not a way of putting off that consultation, and a delayed diagnosis is a real loss, not a theoretical one.
The case that gets cited:
a 22-year-old woman, escitalopram and ashwagandha
Presented in 2025 at the congress of the American Academy of Neurology. A 22-year-old woman had been taking escitalopram 10 mg a day for two years. Two days before admission she took ashwagandha: a 600 mg capsule and a tea. She arrived at hospital with clonus in her limbs, eyelid tremor, tachycardia, fever, vomiting, dilated pupils and cardiac arrhythmia. She was treated by stopping the serotonergic agents, with benzodiazepines and supportive care, and seen by psychiatry and toxicology.
What this case proves: that it can happen. What it does not prove: how often. A case report is the weakest form of evidence in the hierarchy and, at the same time, the only one that exists for many interactions. It is worth reading in exactly that way: not as a ban, but as a description of how things can go wrong. For context, the clinical criteria for serotonin syndrome note that almost half of cases occur with a single serotonergic agent, not with a combination.
The signs that mean A&E, not a forum. The clinical criteria require recent exposure to a serotonergic agent plus at least one of the following: spontaneous clonus; inducible clonus with agitation or sweating; ocular clonus with agitation and sweating; tremor and exaggerated reflexes; rigidity with a temperature above 38 degrees and ocular or inducible clonus. Most cases appear within the first 6 to 24 hours of a change in dose or the introduction of a new substance.
The first window is the shortest and the noisiest. The clinical criteria for serotonin syndrome say most cases appear within the first 6 to 24 hours of a change, and adding a new product on top of an antidepressant is a change. If clonus, tremor with exaggerated reflexes or fever appears, do not wait until morning.
The second report, far less spectacular:
prolonged sleep, confusion, amnesia
Also in 2025, a case report describes a 40-year-old man who saw a doctor with prolonged sleep, abnormal sleep behaviours, acute confusion and retrograde amnesia, appearing in a temporal relationship with taking ashwagandha. A full assessment ruled out other causes. Stopping the supplement, plus supportive care, led to complete resolution of the symptoms. The authors attribute it to GABAergic modulation by withanolides and to individual variability in metabolism.
The detail that matters: that man was not taking any sedative. The plant did this on its own, in a susceptible person. Add a benzodiazepine or a sleeping pill on top of it and you do not need an exotic interaction, you just need arithmetic. The US professional fact sheet lists drowsiness among the plant’s common effects, and the French nutrivigilance dossier includes it among the effects reported by consumers.
Statistically, this is the far more likely version of the problem. Not a textbook emergency, but a day when you are sleepier than you should be, at the wheel or at work. If you want the right order for supplements over the course of the day, we have written separately about which supplements should not be taken together and at what time.
Practical, not theoretical. If you add ashwagandha on top of something that sedates you, the first few days are not the days to test how alert you are at the wheel. No alcohol the same evening, no machinery, no night shifts where you rely on your reflexes. And not for children: Poland has explicitly banned ashwagandha products for children, pregnant women and breastfeeding women.
The overlap nobody talks about:
the liver
The US hepatotoxicity database has a dedicated entry for the plant. Up to 2024 there were around 23 published cases of clinically apparent liver injury. The pattern is fairly consistent: it appears 2 to 12 weeks after starting, and it is usually cholestatic or mixed, with jaundice and itching. The likelihood score assigned is B, meaning “likely cause of clinically apparent liver injury”. Most cases are mild or moderate and resolve within one to four months of stopping, but there are also rare cases that were fatal or required a transplant, mostly with pre-existing cirrhosis.
The US professional fact sheet adds the dose range from the reported cases: 450 to 1,350 mg a day, for one week to four months. And in France, the nutrivigilance dossier of the health agency, published in its opinion of 19 April 2024, contains 8 reports of adverse effects since 2009, of which 6 were assessed for causality and 4 classified as “plausible”, including one acute hepatitis.
The link with the subject of this article is a practical one, not a pharmacological one. Many psychiatric treatments have their own entries in the same database, and the doctor who prescribes them knows which. The problem arises when you add a product on top of them that the doctor does not know about: if your blood tests come back abnormal, there is no longer any way to tell what should be stopped. We discuss the same logic in the article on ashwagandha and the contraceptive pill, where the real overlap was also the liver.
What you never do:
you do not stop the antidepressant to make room for the plant
It is the thought that comes on its own after a couple of good weeks: if the supplement is working, maybe I can cut down the pill. It is also the most dangerous thing in this whole article, and the figures leave no room for interpretation.
| What was measured on stopping | Figure | Source |
|---|---|---|
| Withdrawal symptoms after abrupt stopping, within 5 to 8 days | 60% and 66% | randomised double blind study, sertraline and paroxetine, after about 11 months of treatment |
| Severe withdrawal | 1 in 35 | systematic review, 2024 |
| Moderate or severe withdrawal, treatment for more than 2 years | 64% | London survey, reported in 2025 |
| Onset of symptoms after a missed dose | 12 to 24 hours | the specialist literature cited in the same review |
| Relapse after stopping medication, in the first month | 25% | the introduction to the very study on ashwagandha on top of SSRIs |
| Relapse in the first year | 80% | the same source |
It is worth noticing where the last pair of figures comes from: the introduction to the paper on which the whole “ashwagandha helps on top of SSRIs” argument rests. The same team that showed a benefit on top of medication writes, two paragraphs away, how badly stopping it goes. The study was never an argument for replacement.
Nothing in this article is a reason to stop, reduce or skip a dose of an antidepressant or an anti-anxiety medicine. Reducing, when appropriate, is done in steps and with the doctor who prescribed the treatment. A supplement that seems to help is not a sign that you can give up your medication, it is at most a topic for your next check-up.
Five things people say:
and what survives a careful reading
Not in the sense in which the expression is used. In the European Union the plant has zero authorised health claims, so the wording is not even allowed to appear in a commercial text. And in the 2025 meta-analysis of 14 randomised studies and 713 participants, the effect on anxiety symptoms held up under additional checking, but the effect on depression did not. The authors themselves list their limitations: few studies, high heterogeneity between them, outlier effects.
This is the position of the most read result in Romanian, and the practical conclusion is reasonable. The reason given is not. The only randomised study that asked the question gave extract precisely to people already taking SSRIs, for six weeks, and the reported result was a larger fall in the anxiety score in the extract group. A “no, full stop” that does not mention this study is not caution, it is not reading.
The mechanism the authors of the randomised study point to is not serotonin, it is GABA. The same explanation appears in the 2025 case report of prolonged sleep and confusion, where the authors talk about GABAergic modulation by withanolides. Serotonin is the hypothesis from a single case of serotonin syndrome, not the plant’s demonstrated mechanism. The practical consequence is the reverse of what is written: the theoretical overlap is with the anti-anxiety medicine, not with the antidepressant.
The figures are unforgiving. In a randomised study on stopping, 60 and 66 per cent of participants taking sertraline or paroxetine had withdrawal symptoms after five to eight days. A 2024 review estimates one case of severe withdrawal for every 35 people who stop. And the introduction to the very study this article relies on notes that stopping medication leads to relapse in 25 per cent of cases in the first month and in 80 per cent in the first year.
It is not sold everywhere, and not under the same conditions. Denmark withdrew ashwagandha products from the market after a 2020 risk assessment by the Technical University of Denmark, which could not establish a safe lower limit of intake. Poland capped root powder at 3 g a day and withanolides at 10 mg per recommended daily dose. Being available on the shelf is not a safety assessment.
What you find in the top results in Romanian:
either silence or the wrong mechanism
We read the top results that come up for a search in Romanian. Only one tackles the subject head on.
- natu.care says explicitly that ashwagandha must not be combined with antidepressants, including SSRIs, SNRIs, tricyclics and monoamine oxidase inhibitors, and the reason given is its action “on serotonin receptors”. It even cites the 2021 study of 40 people, but again as an argument for caution. The mechanism the study’s authors point to is a different one.
- catena.ro warns that “interactions may occur with anxiolytic, sedative or antidepressant medicines”, without a single figure, and lists six French wellness websites and supplement shops as its sources.
- smartliving.ro warns about benzodiazepines, anticonvulsants and barbiturates, which is correct, but in the same article says the plant contains tryptophan and raises serotonin, without connecting the two ideas.
- sfatulmedicului.ro does not mention medication at all and says the effects on anxiety are “comparable to those of dedicated medicines, but without side effects”.
- A medical forum thread appears on the first page with the title “strongly antidepressant, anxiolytic and anti-stress effects”, exactly the wording nobody is allowed to use in commercial communication in the European Union.
None of them mentions the escitalopram case report. And none of them says the most useful thing about the study some of them cite: that it was done in people who were already taking SSRIs, which is exactly the situation the reader is asking about.
What is in our gummies
and what we tell you before you take them
The product contains 600 mg of root extract standardised to 5% withanolides, in two gummies, made in Romania, with the formula printed on the label. The plant has been used for centuries in Ayurvedic medicine. We have written separately about dosage and time of day in when to take ashwagandha and what the optimal dose is, and if you have a thyroid problem, first read ashwagandha and the thyroid.
And the thing we say most clearly, because it is the very subject of this article: if you take an antidepressant or an anti-anxiety medicine, we are not selling you this product as a solution. Ask your doctor first. The data that exist are one six-week study of 40 people and two case reports, and that is not a basis for deciding on your own.
The product in this articleAshwagandha KSM-66, in gummies600 mg of root extract standardised to 5% withanolides, two gummies a day. Not intended for people taking psychiatric medication without their doctor’s approval.View the productFrequently asked questions
That is a question for the doctor who prescribed your antidepressant, not for an article. What an article can tell you is what is on the table: a single randomised study, in 40 people already taking an SSRI, in which the ashwagandha group improved more on the anxiety scale, and a case published in 2025 of serotonin syndrome in a 22-year-old woman who had been taking escitalopram for two years. One small study and one case. That is not enough to say yes, or to say no.
Here the overlap is more plausible than with antidepressants, not less. The authors of the only randomised study describe ashwagandha as a plant with a GABAergic property, and benzodiazepines act precisely on the GABA-A receptor. The US professional fact sheet lists sedatives among the medicines the plant may interact with, and drowsiness is a common effect. Two things that sedate, taken on the same evening, sedate more.
No. In the European Union it has no authorised health claim, so nobody is allowed to attribute an effect on mood to it in commercial communication. And in the meta-analysis published in 2025 of 14 randomised studies and 713 participants, the effect on anxiety held up under additional analysis, but the effect on depression did not. The word “antidepressant” in headlines on the internet does not come from the literature.
The clinical criteria used for serotonin syndrome require recent exposure to a serotonergic medicine plus at least one of the following: spontaneous clonus, inducible clonus with agitation or sweating, ocular clonus with agitation and sweating, tremor and exaggerated reflexes, or rigidity with a fever above 38 degrees. Most cases appear within the first 6 to 24 hours of a change. If they appear, you go to A&E, you do not search a forum.
Not on your own, ever. In a randomised study on abrupt stopping, 60 and 66 per cent of people taking sertraline or paroxetine had withdrawal symptoms, and these can appear 12 to 24 hours after a missed dose. A 2024 review estimated one case of severe withdrawal for every 35 people who stop. A dose reduction is done with your doctor, in steps, not because a supplement has appeared in the house.
It means that neither we nor anyone else are allowed to write in a commercial text that it reduces stress, anxiety or cortisol. What is allowed is something else: how the plant is traditionally used, exactly what a product contains, what studies have measured and what they have not. The rest of this article is built on that rule, including the part where we criticise our own product page.
Note: the information in this article is for educational purposes. It does not replace personalised medical advice and does not replace a conversation with the doctor who prescribed your psychiatric treatment. Ashwagandha has no authorised health claims in the European Union, and this article does not attribute any health effect to it. Food supplements are not medicines and are not intended to diagnose, treat or prevent disease. Do not exceed the recommended daily dose. Keep out of the reach of children.
Comments (0)
There are no comments for this article. Be the first one to leave a message!