• Increase font size
  • Default font size
  • Decrease font size
Home Performance Supply Micro Intervals for TT

Micro Intervals for TT

E-mail Print PDF

The 2005 racing season is just around the corner, bringing with it new goals, new motivations, and the training plans to match.  I’d be willing to bet that most of those plans come with some kind of interval training.  These lovely little devils are always lurking around, usually guilting us into putting ourselves at their mercy, all in the hopes of being able to go a little bit faster on our bikes.  You just gotta love ‘em!  I usually do…but only after finishing them off and proclaiming that I whupped those devils once again!


So, what’s your interval poison of choice?  VO2’s?  Lactate threshold?  Anaerobic capacity?  It’s likely a combination of all three, in a variety of forms, whether you do them in training or get your fill of them while racing.  But what about micro-intervals? 

Micro-intervals (MI’s) are short, on-off bursts which allow one to intermittently work at a power output which is much higher than could be performed continuously.  MI’s are sometimes performed in a 15 seconds on, 15 seconds off manner over a longer, continuous duration, although other variations exist.  The high power output during the “on” part of the intervals then accumulates into a significant amount of total, high-power stimulus to one’s body.  Hypothetically, this large amount of high power stimulus could lead to adaptations which supposedly can improve one’s time-trial performance, or at least that is one reasoning behind the practice and prescription of these MI’s.  Since Chris Boardman purportedly used micro-interval’s (MI’s) of some type in preparation for his 1996 “unlimited” Hour Record, I occasionally hear of riders and coaches using MI’s for time-trial preparation, particularly in the weeks preceding an important event.  Are MI’s really productive in practice though?

This fall, I decided to investigate these intervals to see what effect two weeks of MI’s would have on 20 minute maximal power (i.e. a 20 minute time-trial).  I recruited two subjects to punish themselves in this study: myself, who I classify as a “moderately trained” cyclist (see graph below for the Training Stress Score, or TSS) and my brother and BTR grand-master Kraig, who I call an “under-trained” cyclist (0-1 hour/week for MANY preceding months).


 Here were the methods:

  • -Power measured with SRM PCV’s, zeroed each day
    • 20 minute maximal power (20MP) test as a baseline on an indoor trainer
    • 2 weeks (7-8 sessions/days) of MI training
    • 20 minutes of 15 second on, 15 seconds off per session/day
    • “on” @ >150% of 20MP
    • “off” @ “soft-pedal” power
    • ~1 hour total training session duration

no other significant training was performed in order to keep my over-all training load similar to the weeks      preceding the study, to focus the study on MI adaptations, and to speed MI session recovery

  • 20MP test on an indoor trainer after the completion of MI training weeks
  • 20MP tests were performed at ~ the same time of day (morning) with similar diet that morning
  • Cadences were self-selected for tests and MI’s
  • I performed MI sessions both indoors and outdoors
  • Kraig performed MI sessions exclusively indoors
  • Power for tests and MI sessions was normalized per Andy Coggan's protocol

So, off we went for two weeks of beating down those interval demons!  It was an interesting two weeks.  There was plenty of suffering, but also a sense of satisfaction in taking on these intervals and finishing them off.  Did our suffering pay off with us becoming better TT’ers?  Here is the 20MP test data:

As you can see, I had a decrease of 6 watts in my 20MP over the two weeks.  There was no wattage boost for me.  I basically just suffered for nothing, at least in terms of improving my time-trialing.  Kraig, on the other hand, stepped up a bit with 16 more watts!  Of course, I suspect that with as little riding as he had done in the previous months, just the fact that he rode his bike regularly was probably responsible for most, if not all, of his improvement…although, I do have confirmation that he was wearing REDSHOE BRAND shoes (note the similar % gains in 20 minute power during the first two weeks of Project: Redshoe and the MI test as shown below).  Hmmmm.  Maybe I need a pair of those shoes?


Looking at the data and observations from the two-weeks of training, there is plenty more to learn.  First, let’s take a look at the average power output during the MI workouts themselves and the 30 second rolling averages for those workouts:


Purple:     intervals performed indoors

Blue:        intervals performed outdoors


There are a few things which jump out right away.  First, the normalized MI power (MIP) data may point to why I did not improve my 20MP and Kraig did.  My MIP was significantly lower than my 20MP (30 watts on average).  This implies that the MI’s on average were not sufficiently intense enough for long enough to overload the energy systems responsible for generating my 20MP.  Kraig’s MIP was near his first 20MP test and eventually exceeded it.  He likely did overload the systems responsible for his 20MP and that may have resulted in some adaptation. 

The second thing that jumps out is that both of us showed a trend towards increasing MIP outputs.  The systems stimulated during the MI’s apparently adapted to some extent.  In other words, we both got better at MI’s.

A third observation is that the “on” part of our 30 second rolling averages increased during the study, and these values relate very closely to the increases in MIP.   

There was another important observation as well: my rating of perceived exertion (RPE) was completely out of whack with what my MIP would indicate.  In other words, the MI’s hurt much more than a continuous effort at the same normalized power.  I really struggled to complete each MI session.  I could not complete but 6 minutes of MI’s on 10/30 (note the missing data), and I was a bit short of 20 minutes on 11/2 and 11/4.  It felt like I was doing a 20 minute time-trial each day when on average I was going much “easier”.  There was a lot of pain with no gain, at least in my 20MP!  Kraig initially did not have this RPE discrepancy, but it did show up in several of the workouts in the middle of two weeks of MI training (note the missed workout on 11/02 which was due to mental trauma).  Perhaps this was because his MIP was pretty close to, and then exceeded, his first 20MP test.   

Looking deeper into the data, there is more to learn.  If I didn’t improve my time-trialing, what did I improve during those MI’s?  Let’s look at one of my MI sessions and see what we can discover: 

Ouch!  Look at those MI’s.  My heart-rate was similar to what I had for my 20MP tests, but as you can see, at a much lower normalized power.


Power distribution for the MI’s on this day:

Lots of time spent over 400 watts! …and lots under 100 watts.  Hardly any time spent near my 20MP.

Also, here is a plot of my pedal force vs. cadence for a MI session and my concluding 20MP test:

The ~30% greater pedal forces at ~10% faster muscle contraction speeds during the MI’s - the blue data above - (as compared to the 20MP test), lends evidence to suggest that there was significantly different muscle fiber recruitment during the MI efforts.  The data implies that I called upon significantly more fast-twitch muscle fibers during the MI’s than the 20MP tests. 

So, just what power production systems did we train?  Well, considering that I improved my MIP and MI “on” power (just like Kraig) but slid a bit on my 20MP (within the error of the measurement though), it looks like I overloaded the systems used to generate 400+ watts (in 15 second bursts) more than the one’s used in 20MP production. 

I would speculate that the 10 or so minutes of “on” time during the intervals stimulated both MIP (specifically the “on” portion) and 20MP production systems; however, the duration of the “on” was not long enough to overload my primary 20MP production systems in a manner which would carry over to continuous efforts since the MIP was significantly below my 20MP.  The only real conclusion to be made here, is that each MI “on” portion appears to have been sufficient to overload the systems used to generate 400+ watts in 15 second repeated bursts:  hence, the general improvement in MIP over the two weeks.  In addition, the much higher pedal forces and speeds evident during the MI’s hint that muscle fibers not normally used for 20MP production were stressed.  It appears that for me, the adaptations in this “faster twitch” fiber profile did not lead to improvements in my 20MP.

Kraig probably had additional things going on which contributed to his 20MP improvement.  Given his lack of regular exercise preceding the study, he likely boosted his blood-volume rather quickly with the significant increase in activity.  Increased blood volume appears to be accompanied by a rapid rise in VO2max via increased cardiac stroke volume and output.  This short-term adaptation is likely responsible for the majority of his 20MP performance jump.  If Kraig had done a consistent week of training before the study, he probably would not have improved his 20MP as much (if any), or if the study was longer, he probably would not have seen additional improvements of similar magnitude. 

What does all this hard work mean in the end?  I speculate that if I want to improve my 20MP, as a moderately trained (or more) rider who is unlikely to realize an “under-trained” bonus like Kraig, my time is best put towards more continuous effort intervals which better simulate and overload my 20MP systems.  Perhaps by doing MI’s sessions for a longer duration with the “on” or the “off” closer to my 20MP (as opposed to >150% of 20MP) and/or by extending the oscillations to 30 seconds on/off, I could indeed spend a very large amount of time at this lower power level, but the problem with doing MI’s like that is that they still do not necessarily lead to the same adaptations which result from more continuous efforts at the “on” power.  The MI adaptations which would apply to continuous efforts are likely only the equivalent of the average power output for the entire MI session.  Perhaps there were other adaptations taking place which would surface in time beyond the two-week study, but there was no hint of them in the short-term. 

Additionally, MI’s do not seem to be very time-efficient considering that I can generate a higher average power during a continuous interval, or more continuous supra-20MP intervals, within the 20 minutes allocated.  Given this inefficient accumulation of overall training load (TSS) with MI’s, it is unknown if future benefits for the MI’s on 20MP would surface.  It is even plausible that a highly trained cyclist could detrain their 20MP by doing MI’s if the average power is much less than their 20MP!

Future examination may determine if MI’s done in the fashion of this study are effective in improving anything beyond the power one can generate in 15 seconds on, 15 seconds off bursts.  For now though, the only potential race-applicable benefit I see from these MI’s is in simulating the short power surges sometimes needed to position in a pack, to hop on a wheel, or in similar cyclo-cross and mountain-bike efforts.  This case-study does not show MI’s to be effective in creating short term (a couple weeks) time-trialing improvements.

To take a look at MI’s with a slightly different perspective consider this: it has been estimated that I would need to generate something in the neighborhood of 376 watts for an hour to win an Olympic Gold medal in the time-trial.  If I were to use the proposed logic of some MI prescribers, does it seem reasonable to suggest that if I train the ability to do MI’s with the “on” near 376 watts and then progressively build the “on” time in one session up to an hour that I could win Gold?  Heck, I can do well over 400 watts for 10 minutes in one session already!  Three and a half years of work and I’ll be there, right?  Unfortunately, it just doesn’t appear to work quite like that.  In the bigger picture, perhaps the adage that “you have to train that fast to be able to go that fast” doesn’t apply that well to cycling if “that fast” is broken down into too small of pieces.  It is entirely possible that Chris Boardman went as fast as he did in spite of the MI’s he purportedly used (talent has its perks!).  Maybe there is some data out there sitting on someone’s computer which has the potential to dispute this by demonstrating how MI training improved his 1-hour MP, but then again, maybe he was wearing some special shoes under those booties!

More study is certainly needed in order for solid conclusions on MI’s to be made, so if anyone is interested in putting MI’s through the two-week test themselves, out of curiosity or winter boredom, maybe you can help!

Additional Reading 

Billat, L. Veronique. “Interval Training for Performance: A Scientific and Empirical Practice.  Special Recommendations for Middle- and Long-Distance Running.  Part I: Aerobic Interval Training”.  Sports Medicine 2001; 31 (1): 13-19.

Billat, L. Veronique. “Interval Training for Performance: A Scientific and Empirical Practice.  Special Recommendations for Middle- and Long-Distance Running.  Part II: Anaerobic Interval Training”.  Sports Medicine 2001; 31 (2): 75-90.

Billat, L. V., J. Slawinski, V. Bocquet, P. Chassaing, A. Demarle, and J. P.  Koralsztein. “Very Short (15 s – 15 s) Interval-Training Around the Critical Velocity Allows Middle-Aged Runners to Maintain VO2 max for 14 minutes”.  Int J Sports Med 2001; 22: 201-208

Coggan, Andrew. “Training and racing with a power meter:  an introduction”. 25 March 2003. 

Goodman, Jack M., Lui, Peter P., and Howard Green. “Left Ventricular Adaptations Following Short-Term Endurance Training”. Journal of Applied Physiology. 24 September 2004 [online abstract].

About our Contributor:

Kirk Willett, is a twenty-year+ participant in the sport of cycling who has competed in 17 different countries on 5 different continents.  Originally from Pullman, Washington, his racing career has ranged from his roots as a Pacific Northwest junior and amateur competitor to time with the U.S. National Team and then on to professional competition as a member of the Mercury Cycling Team including events such as the Tour of Switzerland.  He was also a director with the Mercury Cycling Team and then directed the Prime Alliance professional team full-time from 2001 through 2003.  He has also been a coach and advisor to members of both the Mercury and Prime Alliance professional teams in addition to other Pacific Northwest athletes. 

Kirk is currently a medical student attending Oregon Health Sciences University building on his exercise science education from Washington State University.  He resides in Portland, Oregon with his wife Tina and two sons.  He is a strong advocate for clean, ethical sport and encourages all athletes to take the same pledge he did as a young amateur:  “I will never participate in doping no matter what I stand to gain.

Last Updated on Monday, 28 December 2009 00:42