Tuesday, March 25, 2008

Effect of Cycling Cadence on Running in a Triathlon

In doing some checking of the research while I was writing a magazine article on triathlon transitions I came across an interesting French study (1). The scientists wanted to see if there was an effect of bike pedal cadence on a subsequent run performance. Eight experienced triathletes completed three bike-run sessions. In each bike portion they rode at 90% of lactate threshold, which is roughly FTP, for 30 minutes. This would have been an effort similar to or slightly harder than what would be done in an Olympic-distance triathlon. They then quickly transitioned to a run to fatigue at 85% of max velocity, about the pace that would be run in a sprint-distance tri. The variable they manipulated was cadence on the bike. In the last 10 minutes of each 30-minute ride the athletes pedaled at 1) a freely chosen cadence, 2) a cadence 20% higher than #1, and 3) a cadence 20% lower than #1. So, for example, if #1 was 90 rpm then #2 was 108 rpm and #3 was 72 rpm.

What they found was that when pedaling at a cadence 20% lower than freely chosen the time to exhaustion on the run increased by 37% on average over the freely-chosen-cadence performance. Running cadence was unchanged by the cycling cadences. They also found that near the end of the low-cadence cycling bout the stress placed on the athletes' as measured by VO2, heart rate, ventilation rate and lactate accumulation were not significantly different compared with the freely chosen cadence bout. The high-cadence bout fared worse than the other two on all measures including time to exhaustion.

The study was interesting for me because it contradicted a study out of the University of Colorado from a few years ago with which I was familiar (2). Thirteen experienced triathletes completed three bike-run sessions on separate days. In each they rode 30 minutes at a high intensity and then ran a 3200-meter time trial. The only thing that changed was the cadence on the bike. As with the French study they rode one at a freely chosen cadence, a second at a cadence 20% higher than freely chosen and a third at a cadence 20% lower than freely chosen--for the entire 30 minutes (this is different than the French study in which only the last 10 minutes on the bike varied). The Buff researchers found that after cycling at a high cadence (100-110 rpm) the run times were nearly one minute faster on average than with the freely chosen bout. And, besides that, the run cadences were also quicker. Stride lengths and observed biomechanics were unchanged.

So there you have it. Take your pick: pedal at a low cadence before entering T2 or pedal at a high cadence before T2. One of them will improve your run performance. Which one? I wish I could say. I have not come across another study on this topic yet. There may be individual differences which affect the results such as your position on the bike, how steady or variable your bike pacing was, how quickly you transition, etc. If you have a personal solution for this dilemma please feel free to post it.

1. Vercruyssen, F et al. 2005. Cadence selection affects metabolic responses during cycling and subsequent running time to fatigue. Br J Sports Med 39(5):267-72.

2. Gottschall JS et al. 2002. The acute effects of prior cycling cadence on running performance and kinematics. Med Sci Sport Exerc (34(9):1518-22.


At March 25, 2008 8:04 AM , Anonymous Rick Kiley said...

Joe, Very interesting post; thank you. My immediate reaction to the study parameters is that +/- 20% is a very large range. In addition, are the sample sizes, given the type of study, fairly small? As such, is it possible that the results are random?

Keep up the great posts. Rick

At March 25, 2008 8:13 AM , Blogger Joe Friel said...

Rick--The sample size for this study is fairly typical of sport science research. Certainly with more subjects the results may have changed slightly, but I doubt if they would have been significantly different.

At March 25, 2008 8:42 AM , Anonymous Joe Reger said...

First, I always applaud guys like you and Noakes for looking to the research to guide protocol. You've done a great job of dispassionately describing the latest numbers.

Second, I'd like to request a blog post about the theory behind run performance after cycling. My strength is the run so this is a big issue for me.

My sense is that personal muscle composition must be a factor. Those with more fast twitch fibers may benefit greatly from lowering cadence at the end because their muscles have "strength in reserve" to handle lower cadences. Guys like me with more slow twitch fibers need to reduce strain on individual fibers because we're closer to our relative strength limits.

How do biomechanics play into cadence choice? Not from an absolute perspective... there are plenty of cycling articles that discuss how to choose your own optimal cadence. I'd like the focus to be on femur length vs. cadence vs. run performance, for example.

Finally, my guess is that distance plays a role. At olympics we may be able to weather more muscular damage while at an Ironman we can't, so need to employ higher cadences.

And I'm sure I'm missing many points. A post outlining the theory of run performance after cycling would be great. Something that describes the main considerations you make when coaching cycling to defend run performance. Or a pointer if you've already done it somewhere.

Thanks again for the great work on this blog, in your books and in your magazine articles.

At March 25, 2008 8:55 AM , Blogger Mike Hardy said...

The only difference in protocol between the two studies lends itself to the testable hypothesis that in a 30-minute cycling effort followed by run to exhaustion, pedaling the first 20 minutes at +20% natural cadence then the last 10 minutes at -20% natural cadence is the fastest.

Interesting that this goes along with a study I just read about varying power output - the fastest subsequent run times were achieved when you varied power output by +10%/-10% every 5 minutes(? or so) and the last -10% was right before the transition. Average power and cycling time was unaffected but the run was way faster.

Combining the two and maybe varying the effort and cadence with a final 5 minutes of lower-cadence / -10% power work before the run may be the most optimal

All speculation of course but if anyone has an exercise lab, it is testable :-)

Love the blog and exercise physiology research even if it doesn't have firm conclusions yet - thanks Joe!

At March 25, 2008 1:56 PM , Blogger Joe Friel said...

Joe R--Yes, it is an interesting topic for me, also, and I have been pulling up all of the research I can find on it. There's more than I expected. We'll see what comes of it.

Just as a side note, one thing I have found in working with elite triathletes is that they use a lower cadence in Ironman-distance races than in shorter distances. In fact, as the distance gets shorter their cadence continues to increase. I don't know that this means it is necessarily "good" or "beneficial", though. Just an interesting observation based on limited data.

At March 26, 2008 7:40 AM , Blogger David said...

Hi Joe
I would like to add my thoughts to this topic. I used to try focus on maintaining a high cadence (-+ 100) on the bike but recently have begun using a slower cadence (-+80)after seeing data from Ironman pro's and the effects have been brilliant for me, not only faster run times but also faster bike splits!! One of the ideas I had about this is that your legs are rotating mass (much like wheels) and hence the less times you rotate them the less distance they travel for the same amount of work (and anything that rotates travels further than you actually ride). Considering they weight at least 15 times your wheels mass the effect could be far more important than wheel weight. There is one draw back I have noticed and that is recovery seems slower (but this could be to racing more intensly being in better position). This makes sense because more intense contractions take longer to recover from much like a gym workout.The time under load is far less in gym but is closer to your maximum output per rep which seem to take longer to recover from. However during racing slower cadence seems to have a net saving of energy because of fewer rotations of the leg mass. I have a few more ideas about this issue and the pacing issue but am busy experimenting with them so might add them later.
Thanks that's it for now

At March 27, 2008 8:35 AM , Anonymous Anonymous said...

Isn't there a significant difference between running to fatigue and running speed over a fixed distance? Maybe the lesson to be learned is that slower cadence allows for longer runs while faster cadence allows for faster runs.

At March 27, 2008 7:26 PM , Anonymous Anonymous said...

I think cadence is way over looked. Ex sergio-marques in hawaii 2 years ago cad 72 marathon 2:43 craig alex spin very high faris is around 80rpm. I think power out put has more to do with cad. You just need to find out what works for you If I spin at 100rpm my heart rate will go up 15bpm compared to a cad aroun80-85
we are all different.

At March 28, 2008 7:57 AM , Anonymous Anonymous said...

Hey Joe,
I read your post on the cadence question, and read both studies. Although you and other people with more experience than me will know more about this, I had the following thoughts - which could be completely crazy.

First, I think the initial Colorado study makes more sense to me given that it is not a time to exhaustion trial, and allowed the subjects to use their own race intensities for both cycling and running. I usually think real endpoints, like speed in time trials, are more realistic and therefore carry more weight than time to exhaustion, which is not how people race (I have the same objection to some of the protein nutrition studies), especially when done on a treadmill.

Outside of the endpoint issue, the studies also differ slightly in population - the French athlete cohort was smaller and they had more experience (minimum 5 yrs vrs. 2, which can make a big difference in endurance races) and were older. They also differ in methodology, which I think is a big point - the French study only changed cadence 10 minutes out from transition while the Colorado study did the entire 30 minute cycling bout at a different cadence. It probably makes more sense by the way people race currently to do the French study, but it's hard to know which methodology is better (to me).

That being said it's hard to throw out the French study since the athletes were working less hard in their final ten minutes despite producing the same amount of power by their physiologic parameters, and that may account for their improved speed. So hard to say whether it's less physiologic work going into transition, or perseveration, or what that is accounting for the change.

I would say a repeat study needs to be done with a larger cohort (and that the French study shouldn't change practice yet). The classic altitude studies by Levine and Stray-Gunderson in the late nineties had 39 athletes in 3 arms; getting closer to that with more realistic endpoints and deciding on 10 vs. 30 minutes for the cadence change would help. Also, it may be worthwhile for someone to do this for short course and long course racing.

Anyway, just my thoughts. I really enjoy your blog - now that I've finished Going Long, although it may increase my non-training procrastination factor exponentially!

At March 28, 2008 4:11 PM , Blogger Steve said...

Joe - Have you ever considered facilitating some larger-scale research?

I imagine plenty of researchers would be willing to develop a simple protocol that could be semi-reliably followed by your audience, many of whom are likely both power-meter owners and enthusiastic test subjects. The opportunity for a huge test population has to be intriguing...

Really enjoying the blog.


At March 28, 2008 10:41 PM , Blogger Joe Friel said...

Steve--There are quite a few. I'm in the process of reading through them as time allows.

At June 23, 2008 3:21 AM , Blogger Krsnendu said...

I tend to agree with anonymous. Speed and time to exhaustion are different things.

It seems that a higher cadence helps with leg speed in short course races, and a low cadence helps staying power for longer distance races.

At March 13, 2009 10:27 AM , Anonymous Anonymous said...

One issue seems to be that the protocols where not the same. The best results where both from the second of three runs?

At November 7, 2009 8:04 AM , Anonymous Anonymous said...

I found this site using [url=http://google.com]google.com[/url] And i want to thank you for your work. You have done really very good site. Great work, great site! Thank you!

Sorry for offtopic

At May 6, 2010 3:15 AM , Anonymous David Hussey said...

It is very nice post. I thing that to investigate the effect of the cycling cadences on a subsequent 3000 m track running performance in well trained triathletes.Triathlon


Post a Comment

<< Home