Saffron and ashwagandha hold up across 31 trials, with one liver-shaped caveat

Saffron and ashwagandha are among the most popular herbal supplements for stress, low mood and poor sleep, and both are sold on the promise of calming an overactive stress response. Until now the trial evidence for each plant has been scattered across dozens of individual studies of varying quality, making it hard to say which conditions they genuinely help and at what strength of evidence. This review set out to pull all of that randomised trial evidence together in one place, across eight different mental health outcomes, and to grade honestly how certain the answers really are.

The authors ran a prospectively registered, PRISMA 2020-compliant sweep of the literature, with the search updated to 31 August 2026. From 494 initial records, 452 remained after duplicates were removed, and 34 studies met the eligibility criteria: 19 on saffron, 14 on ashwagandha and 1 comprehensive review, together comprising 31 randomised controlled trials. Outcomes spanned depression, anxiety, stress and sleep among eight stress-related mental health measures, and the team also assessed each trial's risk of bias and graded the overall certainty of the evidence.

Saffron showed efficacy for depression across 10 trials, anxiety across 9 and sleep across 8, at 15-30 mg per day of a standardised extract. Ashwagandha showed efficacy for anxiety across 10 trials, stress across 8 and sleep across 3, at 30-600 mg per day of a withanolide-standardised extract, with a pooled cortisol reduction of 3.27 µg/dL. Risk of bias was low in only 10 of the 31 trials, with some concerns in most of the rest, and certainty was moderate at best. Both plants had favourable safety profiles overall, although ashwagandha-associated liver toxicity means liver enzyme monitoring is warranted during prolonged use.

This review confirms that the clinical evidence for both plants is more substantial than for most herbal supplements, while also drawing a clear line around what remains unproven. Notably, no randomised trials exist for burnout or post-traumatic stress disorder for either plant, despite mechanistic evidence involving stress hormone axis normalisation, reduced neuroinflammation and effects on mood-related brain chemistry. That mechanistic work is predominantly preclinical and still requires clinical confirmation, and the authors flag burnout and PTSD as priority targets for well-designed future trials.

If you have wondered whether saffron or ashwagandha are pure marketing, this is a reason to take them more seriously than most supplement-aisle claims: dozens of randomised trials point the same way for mood, anxiety, stress and sleep. It is equally a reason to hold the enthusiasm at moderate, because most trials had some quality concerns and the certainty never rises above that level. The liver signal with long-term ashwagandha use is worth knowing about, and anyone drawn to these herbs for burnout or PTSD should know those specific conditions have never been tested in a trial.