Rhodiola rosea: What the Trials Show, and How to Read the Label

Abstract

Rhodiola rosea is sold on a claim about fatigue, and it is one of the botanicals where the extract specification matters more than the plant name. Two grams of a herb standardised one way is not the same product as two grams standardised another.

A systematic review screened 206 articles and found eleven controlled trials. Two of six trials on physical fatigue reported a benefit, three of five on mental fatigue did — and every included study carried either a high or an unclear risk of bias.

Boost contains Rhodiola. This article is our account of what that decision rests on, and it does not rest on much.

1.What Was Actually Reviewed

Sana Ishaque, Larissa Shamseer, Cecilia Bukutu and Sunita Vohra, working through the Complementary and Alternative Research and Education programme at the University of Alberta, searched six databases for randomised and controlled clinical trials of Rhodiola rosea for physical and mental fatigue.

Of 206 articles identified, eleven met the inclusion criteria — ten described as randomised trials and one as a controlled clinical trial. That ratio is itself informative about the literature: most of what is written about this plant is not a trial.

Two reviewers independently screened, extracted data and assessed risk of bias, which is the standard that makes a systematic review worth more than a narrative one.

2.The Results, and the Sentence That Matters More

On physical fatigue in healthy populations, two of six trials reported Rhodiola to be effective. On mental fatigue, three of five randomised trials did.

Roughly half positive, in a small literature. That alone would place this somewhere between promising and unproven.

The reviewers' next sentence is the one that changes the reading: all of the included studies exhibit either a high risk of bias or reporting flaws that hinder assessment of their true validity.

Not most. All eleven. Which means the split between positive and negative results cannot be interpreted, because the studies on both sides are not reliable enough to be counted against each other.

Their conclusion is correspondingly restrained: research regarding efficacy is contradictory, some evidence suggests the herb may help, and methodological flaws limit accurate assessment. They call for a rigorously designed, well-reported trial that minimises bias — which, as of writing, we have not been able to find.

3.Why the Label Specification Is Not a Detail

This is the part a reader can actually use, and it applies well beyond this plant.

Rhodiola root contains several classes of compound. Two are treated as markers: rosavins, a group specific to this species, and salidroside, which is not. Extracts are standardised to declared percentages of both, and the ratio conventionally used in the clinical literature is three per cent rosavins to one per cent salidroside — chosen to approximate the proportions in the root.

An extract standardised only to salidroside, or to a different ratio, is a different preparation. It may work better or worse; the point is that a trial of one tells you very little about the other, and the plant name on the front of a bottle tells you nothing at all.

This is why we regard herbal trials as testing extracts rather than plants. The specification is the intervention.

It also means a shortcut is available to anyone selling one. Cite the trials, supply a differently standardised extract, and the citation is technically about the same species while being about a different product.

4.What the Broader Literature Adds

Reviews published since have not resolved the picture. A 2024 review in the British Journal of Nutrition examining Rhodiola as an adaptogen for exercise performance describes a literature that leans heavily on animal work — rats swimming to exhaustion, rodents on climbing tasks — with human trials remaining small and heterogeneous.

Rodent time-to-exhaustion results are mechanistically interesting and they are not evidence about people. We flag this because adaptogen marketing draws on that literature freely without marking the transition.

The word adaptogen itself deserves a note. It denotes a proposed category of substances that increase resistance to stress non-specifically. It is not a regulatory or pharmacological classification, and nothing follows from a substance being placed in it.

5.What We Would Need

The reviewers named it in 2012 and it has not appeared: one adequately powered, pre-registered, well-reported randomised trial of a specified extract at a specified dose, in a defined population, with fatigue measured by an instrument chosen in advance.

Until then the honest description of Rhodiola is that it has been studied for decades, has a plausible traditional basis and a mechanistic literature, and does not have a trial anybody can lean on.

That is a weaker position than most of what is written about it, including by people who sell it.

Editorial Comment

Boost contains Rhodiola rosea at 150 mg, standardised to three per cent rosavins and one per cent salidrosides. We chose that specification because it is the one the clinical literature used, which is the only defensible reason to choose a specification.

What we are not doing is presenting that as evidence the ingredient works. The literature it points to is eleven trials, all at high or unclear risk of bias, split roughly evenly on whether an effect was found.

There is no authorised health claim for Rhodiola rosea in the European Union, so we make none. A reader who takes from this article that our Rhodiola will reduce their fatigue has taken more from it than it contains.

Fatigue that does not lift with rest is worth investigating rather than supplementing. Low ferritin, thyroid dysfunction, sleep apnoea and depression all present that way, all are identifiable, and finding the cause is worth considerably more than any botanical in this literature.