Cordyceps raised oxygen use at rest. So what?

I formulate functional-mushroom supplements, cordyceps among them, so I have a stake in this paper going well. I also know exactly how its title will get used, which is why it is worth reading slowly.

Cordyceps is marketed as the oxygen mushroom: stamina, endurance, better use of oxygen. This publication already graded that endurance evidence as real but inconsistent. The new trial adds a phrase that fits the marketing almost too well. Here is what it measured.

The trial

Hamidreza Farzan and Maryam Koushkie Jahromi, sports scientists at Shiraz University in Iran, ran a randomized, double-blind, placebo-controlled crossover with a 7-day washout. Twelve recreationally active young men, mean age 21.6, took either 1 g of a standardized Cordyceps militaris fruiting-body extract or a cornstarch placebo, 30 minutes before a maximal cycling test to exhaustion. The extract was a commercial product, Real Mushrooms RM-CM150C, standardized to at least 150 mg/g beta-glucans and at least 2.5 mg/g cordycepin. The paper states no specific funding and no competing interests.

Before and 3 minutes after the ride, they measured reaction time on a 25-trial Stroop task, plus blood glucose, blood pressure and heart rate. Oxygen uptake (VO₂) was measured at rest and at peak exercise. The design is sound for what it is. The paper's own limitations section is direct about the size: "a small, homogenous sample (N = 12, recreationally active young men) limits the generalizability of the findings."

What resting oxygen uptake actually is

VO₂ is how much oxygen your body pulls out of the air per minute. At peak exercise, it is a ceiling: the most oxygen your heart, lungs, blood and muscle can move and use. That is the number endurance people care about, and it is what the cordyceps story is really selling.

At rest, the same measurement means something different. Nobody is pushing against a ceiling while seated. The body takes up as much oxygen as its tissues are burning, which is why resting oxygen uptake is how resting metabolic rate is measured by indirect calorimetry. A 2005 Journal of Applied Physiology paper by Byrne and colleagues, which measured it in 769 adults, describes the whole MET system as "expressing energy cost of physical activities as multiples of resting metabolic rate."

So a higher resting VO₂ reads, first and most simply, as a body spending more energy while sitting in a chair. That can have many causes. None of them is automatically a better engine.

The numbers

Resting VO₂ was 4.06 ± 1.68 mL·kg⁻¹·min⁻¹ after cordyceps and 2.80 ± 0.93 after placebo, a difference of 1.26 (95% CI 0.27 to 2.25, p = 0.022). That is about 45 percent more oxygen per minute at rest, by my arithmetic. Note the spread in the cordyceps arm: a standard deviation of 1.68 on a mean of 4.06.

For scale, the textbook value for one MET is 3.5 mL·kg⁻¹·min⁻¹. Byrne's group found that convention too high: their measured average at rest was 2.6 ± 0.4, and they concluded that 3.5 "overestimates the actual resting VO2 value on average by 35%." The placebo value in this trial sits close to that measured average. The cordyceps value sits above even the textbook figure. Byrne measured people under a ventilated hood; this trial measured seated participants after 10 minutes' rest, two hours after a standardized breakfast of about 650 kcal. Different setups, so treat the comparison as a sense of scale, not a diagnosis.

Then the number that measures capacity. Peak VO₂ was 34.75 after cordyceps and 33.82 after placebo, p = 0.490. No difference. Heart rate, blood pressure and blood glucose did not differ between conditions either.

Put those together and the physiological finding is narrow: one dose moved oxygen use at rest and did not move oxygen use at maximum effort.

The authors' mechanism points two ways

The discussion offers an explanation, and it is worth reading closely. The authors write that the elevated resting VO₂ "likely reflects transient metabolic priming—enhanced mitochondrial oxidative efficiency at the tissue level." In the same passage they propose cordycepin works by "promoting mitochondrial substrate oxidation and proton leak."

Those two ideas pull against each other. Proton leak is the part of mitochondrial oxygen use that does not get coupled to making ATP. A standard review in Essays in Biochemistry puts it plainly: proton and electron leak "have a major impact on mitochondrial coupling efficiency," and the physiological role of the leak through UCP1 in brown fat "is heat production." More leak means more oxygen burned per unit of useful energy, which is the opposite of efficiency.

I am not saying cordycepin causes proton leak in people. The trial measured nothing at the mitochondrial level, and the authors say so: "direct mechanistic evidence... is absent." My point is smaller. Even the paper's own proposed mechanism does not support reading more oxygen at rest as better oxygen use. The earlier section of the discussion goes further than the data, calling the result "suggesting improved oxygen extraction and utilization at the tissue level." A peer reviewer asked for red blood cell measures to test the oxygen-carrying idea; they were not collected, and the paper lists that as a limitation.

The reaction-time result, three ways

The other half of the title is faster reactions, and here the statistics are more interesting than the headline.

The paper reports a significant main effect of condition (p = 0.040) and no significant interaction between condition and time (p = 0.337). In plain terms, the cordyceps sessions were faster overall, and exercise did not change that gap. The authors say this themselves: the effect "represents baseline cognitive elevation rather than modulation of exercise-induced changes." So a benefit specific to the exhausted state is not what this trial found. The workout made both conditions faster, which the authors attribute likely to practice effects and arousal.

The pre-exercise gap was actually the larger one: 1,183.33 ms on placebo versus 1,048.75 ms on cordyceps, a 134.58 ms difference that was not significant (95% CI −279.44 to 10.28, p = 0.167). The post-exercise gap was smaller and did reach significance.

And the headline post-exercise number appears in three versions in the same paper:

• Abstract: 963.92 ± 84.99 ms versus 1,034.33 ± 84.16 ms, d = 0.88.
• Results text and Table 1: 964.08 ± 78.63 versus 1,042.83 ± 79.52, a mean difference of −78.75 ms, d = 0.83.
• Discussion: "a 70.42 ms advantage at post-exercise assessment, representing a 6.8% performance improvement."

The pre-to-post changes differ too. The abstract gives −148.92 ms for placebo and −104.67 ms for cordyceps. The table means work out to −140.50 and −84.67, by my arithmetic. A peer reviewer flagged that numbers in the text did not appear "fully consistent" with a figure, though that comment was about the oxygen figure, not this table. None of these differences flips the direction of the result, and the raw data file is posted with the paper. I did not re-run it. But when the headline effect of a 12-person trial has three published sizes, the honest reading is roughly 70 to 80 ms, with a confidence interval whose lower end sits near zero. The authors' own response to reviewers gives that interval for the effect size: d = 0.83, 95% CI 0.01 to 1.65.

What the paper says about itself

To the authors' credit, the practical-implications section is careful where it counts: "replication in larger samples is needed before clinical or practical recommendations can be made." In their response to reviewers, they wrote that they had tempered the abstract and conclusion and "now consistently describe the results as 'preliminary,' 'hypothesis-generating,'" and as requiring replication. In the published version, the word "preliminary" appears once in the body, and not in the abstract or the conclusion. The abstract is what most people will read.

The same section also suggests the 70 ms effect is "comparable in magnitude" to reaction-time differences linked to sport performance, and floats e-sports as a use while noting "no direct evidence yet supports this application." One reviewer called the sports-performance link "somewhat speculative." I agree with the reviewer.

How to read the claim when it shows up on a label

The wordings below are my own examples of how this finding could be phrased in marketing, not quotes from any product.

Increases oxygen uptake. Ask: at rest or at peak? In this trial, only at rest, and that is a measure of spending, not capacity. The capacity number, peak VO₂, did not change after one dose.

Faster reaction time after exercise. The trial found a general difference across both timepoints, not an exercise-specific benefit, in 12 young men, on a 25-trial test, with the size reported three ways.

Clinically studied dose. This was one 1 g dose of one specific extract with a stated cordycepin and beta-glucan content. A product with a different extract, a different standard, or no standard at all was not in this study. That is the same point this publication made about lion's mane: a trial is evidence about the material that was tested.

This is a real, blinded, placebo-controlled human trial on a mushroom with very few of them, and it deserves a larger follow-up. It is a reason to run that follow-up. It is not a reason to believe cordyceps gives you more oxygen.

Not medical advice. This is educational analysis, not a recommendation — a study is not a prescription. Talk to a qualified clinician before acting on anything you read here. Full disclaimer →

Disclosure, plainly: I founded and run Shroombiosis (a company I run), which formulates and sells functional-mushroom supplements, including cordyceps. That is a direct commercial stake in the category this report grades, and the extract tested here is a competitor's product. Nothing here is sponsored, no link earns a commission, and I have no relationship with the authors or with Real Mushrooms; here's the full policy. A recommendation with no stake at all: for performance nutrition, Die Tryin Co. is a fellow combat-veteran-owned brand I'm glad to point people to. I don't own it and earn nothing from the link.