Ashwagandha: what the evidence actually supports

Ashwagandha (Withania somnifera) has moved from a traditional Ayurvedic ingredient to one of the most heavily marketed supplements in the wellness aisle. That popularity has outrun the evidence in places, so this guide separates what is reasonably supported from what is being claimed.

What ashwagandha is

It is a shrub whose root and leaf extracts have long been used in Ayurvedic practice, a traditional system of medicine originating in India. NCCIH’s overview of Ayurvedic medicine describes that system and notes that its remedies are not held to the same evidence standards as conventional medicines. Modern supplements are typically standardised root extracts, often sold with a stated percentage of withanolides — the compounds most commonly measured, though not proven to be the sole active fraction.

What the evidence supports

NCCIH summarises the human research. Two areas have the most support, and both are qualified:

• Sleep and stress. Some ashwagandha preparations may be effective for insomnia and stress. “Some preparations” is doing real work in that sentence — extracts differ, and a result from one standardised extract does not automatically transfer to another product.
• Male fertility markers. Limited evidence suggests that taking ashwagandha for two to four months may increase testosterone levels and improve sperm quality. This is a measured-marker finding, not a demonstrated fertility outcome.

What the evidence does not establish

NCCIH is explicit that evidence is insufficient to support ashwagandha for asthma, athletic performance, cognitive function, diabetes, menopause, female infertility or COVID-19. Its effect on anxiety specifically is described as unclear — which is a different statement from “it does not work”, and also a different statement from the confident anxiety claims common in product marketing.

No published guidance from these bodies supports ashwagandha as a treatment for any diagnosed condition.

Safety, cautions and interactions

Short-term use, up to around three months, appears relatively safe in the studies NCCIH reviewed. Reported side effects include drowsiness, stomach upset, diarrhoea and vomiting. Longer-term safety has not been established.

The more serious signal is liver injury. Rare but documented cases link ashwagandha supplements to liver damage, and LiverTox, the NIH database of drug- and supplement-related liver injury catalogues these reports in detail. Anyone who develops unusual fatigue, nausea, loss of appetite, dark urine or yellowing of the skin or eyes while taking it should stop and seek medical advice.

NCCIH advises avoiding ashwagandha if you:

• are pregnant or breastfeeding;
• have an autoimmune or thyroid disorder;
• have surgery coming up;
• have hormone-sensitive prostate cancer.

Potential interactions are recorded with medicines for diabetes, blood pressure, immune suppression, seizures and thyroid conditions. If you take any of those, this is a conversation for your prescriber rather than a decision to make at the shelf.

How to read the label before you buy

Most disappointment with supplements starts at the label rather than the ingredient. Four checks do most of the work:

• Elemental amount, not compound weight. “500 mg magnesium citrate” and “500 mg elemental magnesium” are very different quantities. The elemental figure is the one that matters, and a label that does not state it is telling you less than it appears to.
• Serving size against capsule count. A bottle of 120 capsules at three capsules per serving is a 40-day supply, not a four-month one.
• The full ingredient list, including excipients. Bulking agents, anti-caking agents and capsule material matter to anyone avoiding gelatin, lactose or specific allergens.
• What the label itself claims. In South Africa, complementary medicines are regulated by SAHPRA. A claim printed on the pack sits inside that framework; a claim added by a retailer's marketing copy does not.

How PowderedPurple handles this topic

We separate three things that supplement marketing routinely blends together: what a product's own label states, what the published research base actually supports, and what we think as editors. Those are labelled differently throughout this guide, and where the evidence is thin we say so rather than reaching for a softer verb.

We do not publish a product record until its retailer source, image rights, label claims and affiliate offer have each been checked and dated. That is why this guide discusses an ingredient without recommending a specific bottle: the ingredient evidence is public and citable, whereas a product recommendation additionally requires verification work we will not skip.