Ashwagandha: What the Evidence Actually Says
Promising but unproven, and here's specifically why.
Ashwagandha is getting a lot of hype on the internet right now. Let’s actually summarize the evidence. What’s strong, what’s weak, and the part you won’t hear much about.
We'll start with the good news. The strongest evidence is for stress and anxiety. Reviews pooling well over a thousand adults found a real, consistent drop in stress and anxiety symptoms, and a dose–response meta-analysis of 22 trials reached the same conclusion. Sleep sits just behind it. So if this herb has a sweet spot, it's calming the nervous system and helping you wind down. And honestly, I think that's what most people go to it for.
But here’s the part the ads skip. These reviews are careful work. They’re good. I will say that. The problem sits upstream, in the trials they pull from, and it’s a science thing across every field. Studies that find an effect tend to get published while studies that come up empty tend to die in a drawer. Publishing houses even contributes to that. They want headline grabbers. So the published record almost always leans more positive than it really is. You just get that tilt, so I found analyses where reviewers tested for it, flagged it, and took it into consideration, to their credit:
Now here's what's actually particular to ashwagandha. The trials show publication bias, with the certainty of evidence rated low to very low. The trials are also small and short, and most of them run on a few branded extracts, often with the maker supplying the product or the funding. And if you stack that field-wide publishing bias on top of this, the honest read is probably: early, small, positive-leaning evidence that usually shrinks as bigger, stricter trials show up. The data we have now kind of tells us which way the needle points. We just don’t know how far.
Now, some of the details. There is a hormone angle. A small trial in older, overweight men found ashwagandha raised testosterone about 15% and DHEA about 18% compared to placebo. That's significant. But the study was funded, and the product supplied, by the ashwaganda manufacturer. I wouldn't necessarily throw the trial out, but you should know about that conflict of interest.
On strength, I think there's a real signal. Ashwagandha plus resistance training beat placebo for muscle, strength, and recovery, apart from donated product, this was not industry funded. Another independent review with no industry funding found a clear endurance benefit, improving VO2max.
A note on safety. Ashwagandha is a rare cause of liver injury, usually reversing once you stop.. So talk to your doctor or just skip it if your liver's compromised.
Bottom line: promising for stress, anxiety, sleep, and maybe some strength and endurance benefit. Real, but modest, and thinner than the hype. Now you can decide for yourself.
References
Each link in script text sits on the specific claim it supports. Full detail, with funding and conflict-of-interest notes, is below.
- Stress and anxiety (strongest evidence). Marchi 2025, *BJPsych Open*, DOI 10.1192/bjo.2025.10885 — significant reduction in anxiety and stress; anxiety effect held under Bayesian re-analysis. No outside funding, none declared. Alsanie/Askarpour 2026, *Complementary Therapies in Medicine*, DOI 10.1016/j.ctim.2026.103325 — dose–response meta-analysis of 22 RCTs (1,391 adults); significant reductions in stress, anxiety, depression. No commercial funding, no competing interests.
- Sleep (next-strongest). Cheah 2021, *PLOS ONE*, DOI 10.1371/journal.pone.0257843 — significant sleep improvement across 5 RCTs. Verified unfunded, no competing interests.
- Publication bias is field-wide, not unique to ashwagandha. Rosenthal 1979, *Psychological Bulletin*, DOI 10.1037/0033-2909.86.3.638 — the original "file-drawer problem." Turner et al. 2008, *NEJM*, DOI 10.1056/NEJMsa065779 — across FDA-registered antidepressant trials, positive results were largely published and negative ones largely not, inflating apparent efficacy. Background context.
- Ashwagandha's own reviews tested for and confirmed publication bias. Marchi 2025, DOI 10.1192/bjo.2025.10885 (Egger p<0.001; GRADE low to very low). Alsanie/Askarpour 2026, DOI 10.1016/j.ctim.2026.103325 (confirmed for stress, anxiety, depression; GRADE very low).
- Testosterone and DHEA, with full disclosure. Lopresti 2019, *American Journal of Men's Health*, DOI 10.1177/1557988319835985 — DHEA-S +18% and testosterone +14.7% vs placebo. COI: funded by, and Shoden extract supplied by, Arjuna Natural Ltd; sponsor declared not involved in design, analysis, or writing.
- Strength and recovery. Wankhede 2015, *Journal of the International Society of Sports Nutrition*, DOI 10.1186/s12970-015-0104-9 — 600 mg/day KSM-66 with resistance training: greater muscle strength, size, testosterone, and recovery vs placebo. COI: extract supplied by Ixoreal BioMed (the maker); authors declared no conflict.
- Independent endurance corroboration. Jayawardena 2025, *Turkish Journal of Sports Medicine*, DOI 10.47447/tjsm.0862 — COI-clean review; improved VO2max by 4.09 ml/min/kg. Note: pools VO2max on a small Indian-trial base.
- Liver safety. NIH LiverTox, Bookshelf NBK548536 — ashwagandha a likely cause of rare, idiosyncratic liver injury, usually resolving within 1–4 months of stopping; severe cases concentrated in preexisting liver disease.