Thursday, July 13, 2017

Cycling/Endurance exercise and O2 sensor use

The main marketing material of both the Moxy and BSX oxygen sensors revolve around improving and monitoring your aerobic fitness and performance.  
Claims from the Moxy site:
  1. Determine Muscle Oxygen Zones
  2. Identify Oxygen Delivery or Oxygen Utilization Limiters
  3. Precision Guidance of High Intensity Interval Training
  4. Manage Long Slow Distance Intensity to Maximize Recovery Effectiveness
  5. Calibration of the Athlete’s Rating of Perceived Exertion

Not to be a spoil sport and critic, but I don’t agree.  That is not to say that I don’t love to look and analyze my saturation/hemoglobin changes.  Believe me, I do.  After coming back from a hard ride, I’m at the computer looking at my watts, sats, heart rate within a few minutes. But has this helped me become a “better” cyclist?  No or at least not yet.   I would like to go over a few thoughts and look at the published data to explain my feelings.

First off, one of the claimed uses of the oxygen sensor tech is measuring your lactate threshold. This is closely related to terms such as MLSS- maximal lactate steady state, respiratory break point, critical power etc.  Basically it is a level of watt output/power that can be sustained for 30 minutes or so.  There have been countless papers on estimating this as well as VO2 max and I will not go into detail here.  Suffice it to say that yes, study data shows the O2 sensor curves can be useful.  But, so is the 30 min time trial, lactate testing strips and 3 min all out test and so on (you could make your own equivalent as well that works for you to follow longitudinally - your best 1, 3, 5 min effort).  So is it worth it, considering the cost and trouble to use?

Another talked about performance enhancing feature (especially by Moxy) is guidance in proper training zones and identify your "limiters" (low cardiac output, poor cellular energy fluxes, mitochondrial function, poor local perfusion-capillary density).  I would say that most endurance athletes could use a boost in all areas (I know I could).  In fact, there is very little published(maybe nothing) about this style of training approach.  Most of what you see published are variations of HIT, SIT training as well as the continued controversy of strength training and endurance.  If there is a paper out there using O2 sensor patterns to create better training routines I would love to see it.
In regards to training zones, there is already good track records of using power zones (Watts), heart rate, and I am unaware of any peer reviewed study showing the superiority of O2 predicted zones training. Certainly for the money, a power meter is hands down the best bang for the buck (even more so than heart rate IMO) in working with training zones as well as performance tracking.

The performance limiter prediction is much more complex to discuss.  In order to understand limitation in reaching a maximal effort a very good understanding of the cardiovascular and cellular energy pathways involved needs to be understood and considered. Not only that, but the hypothetical “limiter” pathway may not always be intuitive.  For example in heart failure, the low cardiac output would be guessed at as the performance bottleneck, but it is not.  Instead, it appears to be due to poor muscular energy utilization and changes in the endurance properties of the muscle.  Therefore labeling a given parameter as your "limiter" is not easy.  Remember, we are not saying you are a poor sprinter (low peak power), bad climber (power to weight ratio) which are easy to understand.  The Moxy forum is bursting with talk of limiters, interpreting desat curves.  Again, I have not found anything in Medline published on this type of investigation except in pathological states (CHF, COPD and pulmonary fibrosis, peripheral vascular disease etc)
Another problem with analysis of limiter data is the consistency and accuracy of results.  I would break this down to several sub factors, namely is the sensor itself giving valid, similar results test to test, and could other factors (temperature, location) change the sensor data making interpretation misleading?
It does appear that the BSX and Moxy are reasonably accurate and consistent but in a Moxy study the error rate went way up at higher intensity (just where we need it the most).  Second (and most significant in my opinion) is the effect of sensor placement.  Even small differences in location on the muscle will potentially give very different readings.  Lets look at this study carefully to really appreciate the variation possible.  A couple of excepts follow:
Here is the sensor placement positions on the quadriceps

This is the important part, look at all the different types of desaturation curves depending on placement (in one subject):

So it would be hard to look at saturation curve shape, pattern and absolute values on different days (very hard to get the exact same spot each time).  In addition factors such as hydration, ambient temperature and skin thickness will all potentially muddy the patterns.

Hypothetically, let's say you did know your particular "limiter", could you target training based on the O2 sensor to change it?  I think this is problematic as well.  Given the degree of variance, using the sensor as a type of power meter would be misleading.  Monitoring your progress would also be subject to the same data "smear".  The great advantage of using your power meter is 1% accuracy on a day to day basis.  There is no way an O2 sensor can do that.  Remember, we are looking for subtle changes, not the 20% boost a novice may see in several weeks.

The BSX sensor company is leveraging the hydration change property into a new product, alerting you when you become dehydrated.  Unfortunately the sensitivity of the hydration change in regards to O2 sat become less sensitive at high temperatures (again, just when you need it the most).  The device appears to be wrist worn, also somewhat of a problem since the wrist is notorious for poor optical heart rate accuracy at high intensity and motion(similar technology).

What about looking at the O2 sats and patterns (if they are reliable enough) to help with fatigue and over reaching?  That alone would be of value, and I am still looking (and wishing) for my data to show something.

Back to O2 desat patterns:
One of my favorite studies looked at the changes in deoxygenation patterns with sensor placement in the leg as shown in the above study and graphics.  The curve patterns were pretty much all over the place.


  So take your pick as to what your “limiter” is.  In addition, as far as I can see there are no peer reviewed, published guidelines for interpreting these curves, let alone how using this data in conjunction with a personalized training program (based on the testing) can make you a better athlete.

Despite the above, I have been fascinated with looking at the data and trying to glean anything of value.  So here are a few efforts of the same short climb(same hill), weekly timing (friday’s), same power (+- 5%), roughly same temp (warm to hot).  The yellow line is the O2 sat, red is the power.

As one can see, depending on the muscle (even the same muscle changes, the L rectus femoris was shown from 2 different weeks), the absolute desaturation, pattern of O2 drop, change is quite different (I left off the total Hb changes which are also different muscle to muscle).  Even the same muscle group will have some variation making statistical analysis problematic to say the least.  Yes, I think it is interesting physiology and great for research but from a practical standpoint, not very helpful.  Until that research is done, published and critiqued, the above technology is ambiguous at best in guiding/monitoring endurance training.

The time, money and effort would be best spent otherwise (Powercranks, optimal nutrition, beet root juice, power meter, interval training of specific types(too many studies to link-maybe another post in the future), aero positioning, pacing strategy, inspiratory muscle training etc).

I do commend the above companies for making quality, affordable products and hope that research will better guide their usage. 

Take away:
  • The major marketed use of muscle oxygen sensor tech in sports is endurance exercise.
  • The data is perhaps interesting but can be highly variable due to sensor placement, hydration, temperature and perhaps other unaccounted for variables.
  • Using this to identify limiters is not demonstrated in the peer reviewed literature.
  • Even if you knew the limiter, training to improve this has also not been specifically looked at in the literature.
  • There are better ways to identify training zones.

1 comment:

  1. Not quite the same, but any experience or opinions with Pnoe?