The beet goes on: nitrates improve cycling time trial performance


As of September 2017, new Sweat Science columns are being published at Check out my bestselling new book on the science of endurance, ENDURE: Mind, Body, and the Curiously Elastic Limits of Human Performance, published in February 2018 with a foreword by Malcolm Gladwell.

- Alex Hutchinson (@sweatscience)


Wow. This study is really impressive. Drinking 500 mL of beet juice 2.5 hours before a cycling time trial improves 4 km TT time by 2.8% and 10 mile TT time by 2.7%. On the one hand, this shouldn’t be surprising, because there have been a bunch of recent studies showing beet juice boosting time-to-exhaustion and reducing oxygen cost… But still, those types of studies often don’t end up translating into real differences in the parameter that matters: actual performance. So, as I said: Wow.

The study, published online last week in Medicine & Science in Sports & Exercise, is legit. It’s from Andrew Jones’s group in Exeter. It’s a properly designed placebo-controlled crossover study. The placebo was beet juice with the nitrates (the active ingredient) filtered out with an ion resin, thus indistinguishable from the active beet juice. The subjects (nine competitive cyclists) visited the lab at least five times before the actual experiment even started, to practice taking the time trials until they achieved repeatability of less than 1%.

To reiterate what’s most striking:

  • Performance benefits of 2.8% (6.26 vs 6.45 min) over 4 km and 2.7% (26.9 vs 27.7 min) over 10 miles.
  • This improvement was achieved with just one dose of 500 mL of beet juice, taken 2.5 hours before the event. (Note that this dose is equivalent to 1.6 kg of spinach or 3.1 kg of lettuce!)

They also took other measurements: the amount of oxygen used was the same with and without nitrates, but the power generated was higher with nitrates. Also, plasma level of nitrites was higher after the beet juice, consistent with previous studies suggesting that the beet juice works because nitrates are converted to nitrites then to nitric oxide, which lowers the oxygen cost of muscle contractions.

On another note (further to this previous post):

The subjects also abstained from using antibacterial mouthwash and chewing gum during the supplementation periods since these are known to eradicate the oral bacteria which are necessary for the conversion of nitrate to nitrite.

So what more is there to say, other than “Buy stocks in beet juice companies pronto”? Well, one caveat is that it hasn’t yet been shown that these results can be duplicated in elite athletes. It’s notoriously easier to produce big improvements in less-trained athletes, and these subjects were recreationally competitive. So further studies will be required in elites. But it’s time to acknowledge once again that my initial predictions when I first heard about this research in August 2009 were wrong, wrong, wrong!

24 Replies to “The beet goes on: nitrates improve cycling time trial performance”

  1. I’m seriously confused. Nitrates in meats are said to be bad, but nitrates for training are good for performance? But are they bad for your health?

  2. Really interesting stuff. Who wants to place bets on how long it will take for beet juice to become the next “super food”?

  3. Relaying a question from a friend:

    “From the abstract (which is all I can get access to):

    BR significantly increased mean PO during the 4 km (PL: 279 ± 51 vs. BR: 292 ± 44 W; P<0.05) and 16.1 km TT (PL: 233 ± 43 vs. BR: 247 ± 44 W; P<0.01).

    "OK, so the mean power output over 10 miles was 233W +/- 43W for the placebo group and 247W +/- 44W for the study group. The total study size was 9 cyclists. How the hell can that correspond to rejecting the null hypothesis of no difference between the groups at the 1% significance level? The mean values for the two groups are 1/3 of a standard deviation apart, and the sample size is 4-5 per group. I don’t need to get my student-t tables out to know that’s not proving anything at a 1% significance level."

    My guess in reply to him:

    "In the jargon of these papers, the bounds are often at the confidence level, rather than at 1sd. At the 1% level, the sd implied by the bounds would be much less than 14. That’s just a guess, but think about it: what exactly was the distribution of power outputs over 4-5 cyclists that would imply a sample sd of 44?"

    What's the real answer?

  4. @Ginny: Good question, and not one I have a solid answer to. Here’s what the paper says: “Finally, it should be recognized that the relationship between dietary nitrate intake and human health is controversial (15,36). While increased consumption of nitrate-rich natural vegetable products is unlikely to be harmful to health (15,36), presently little is known about the effects of chronic ingestion of nitrate at high doses.”

    There’s also an interesting discussion of this question here: This article generally argues against the supposed link between nitrates and cancer, but does say: “The one exception might be when nitrite is consumed directly in processed and cured meats. A recent study has shown that nitrite, when consumed with vitamin C (also added to processed meats) and fat (almost always present in these meats), can increase the production of certain types of carcinogens by 140-fold (9).”

    @x-ine: Like I said, buy your stocks in beet juice (or nitrate supplement) companies now! 🙂

  5. @Phil: Good question, thanks for passing it on.

    Standard deviation is pretty much irrelevant here: all it tells us is that there was a mix of fast and slow cyclists in the group. (As an aside, the sample group is 9, not 4-5: all cyclists did four test rides, two at each distance, in a randomized counterbalanced order.) But we’re not interested in whether Cyclist A’s placebo time was faster (or produced more power) than Cyclist B’s beet juice time. We’re only interested in how Cyclist A’s two performances (at each distance) compare, and how Cyclist B’s two performances compare. So the statistical test they used is the Student’s paired t-test.

    To make an analogy, let’s say I build a machine to subtract two from any arbitrary input number. To test it, I input a range of numbers between 10 and 100,000, and measure the outputs. The standard deviation of my input numbers is going to be enormous, maybe around 30,000; same with the standard deviation of my output numbers. The difference between the two mean values (if I’ve built my machine correctly) will be 2 — an infinitesimal fraction of the standard deviations. But the Student’s paired t-test will show that this difference is significant to an arbitrarily small p-value, depending on how many tests I run.

    In the paper, they also run another type of stats to evaluate the “true effect” of the intervention. This concludes that the beet juice was 98.9 to 99.9% likely to be “beneficial or substantially positive” to the athlete. I’ll send you a copy of the paper in case your friend is interested.

  6. Hello,

    So how do I get my pre-race Nitrate fix without having to drink beat root juice. Your articles last year inspired to attempt to load up on beat juice over a 7 day period last year and I only made through 10% of day one. Beat root juice is disgusting…my apologies to beat farmers.


  7. Thanks, Alex. The paired testing idea occurred to me after I posted. Seems obvious in retrospect, but I’m not always the sharpest knife in the block 🙂

  8. Greg Lehman :
    So how do I get my pre-race Nitrate fix without having to drink beat root juice.

    Look for food-grade Sodium Nitrite (e.g. on amazon) and dose very, very carefully. Keep in mind that a *lethal* dose is only 1g for an average (50-60 kg) adult. Sodium Nitrite is what Nitrate is first broken into in your body. Let us know how it works out.

    [ALEX’S NOTE: No, please don’t do this. As pointed out, nitrite is potentially lethal.]

  9. Thanks for pointing that out, George. Yes, with all due respect to @skyskier, PLEASE DON’T TAKE HIS ADVICE!

  10. “Let us know how it works out” was firmly tong in cheek. Doing sodium nitrite is “Darwin Award” territory.

  11. Also in a few studies they say they took beet juice for 6 days. can the same be done with sodium nitrate to get the same results or would that be harmful to you?

  12. I have tried this just for endurance for a casual workout. I’m not a super athlete, but like to push myself to see if I can get more from a workout. After taking the beet juice for just 3 days, I saw an improvement. My biggest improvement seemed to be in lung function. I was able to breathe a little easier. I have been trying to take about 4 ounces, or 1/2 cup per day. I didn’t juice the roots, though. I thought the taste was okay.

Comments are closed.