Wednesday, November 28, 2018

Training the Diaphram - Using the Hexoskin as biofeedback

In a previous post, the concept of respiratory muscle training was reviewed.  One specific target of this type of exercise is the main muscle of respiration, the diaphragm.  Although as seen in some previous study data, the diaphragm is targeted by "non specific" RMT, it is not singled out as a primary goal of that mode of exercise.  So although we will get some involvement of the diaphragm muscle, it may not be an optimal exercise for it in particular.  Since the diaphragm is responsible for most of the muscle force involved with inspiration, it does seem reasonable to work on this process in a more focused approach.  Other reasons for better targeting of the diaphragm relate to individuals with weak diaphragmatic function either as a consequence of injury, surgery or other medical issues.  Conversely, it may be helpful to avoid taxing the accessory muscles of the neck (using generalized RMT) in folks with neck injury or cervical spine disease by minimizing straining with conventional RMT.

Before getting into my interesting technique to better isolate the diaphragm, let's review an interesting paper that did attempt to better target the diaphragm with RMT.  This study aimed to look at EMG of the various muscles along with pressure dynamics in subjects choosing either standard RMT or a "diaphragm targeted" technique.  The RMT device was a Powerbreathe K3 (I have the K4 which will be shown in my data below).
The breathing technique was described as follows:

I have tried this myself and it is not easy!  It is harder than it looks and ideally you really need three hands - one to hold the Powerbreathe, one for the abdomen and another for the chest motion.  In reality the compromise is to remember how to breathe with the diaphragm and just use your hand for the RMT device.

The EMG data does show a difference in diaphragm vs accessory/chest muscle usage
The positions of
the electrodes were as follows: EMGscm was placed at the
midpoint along the long axis of the sternocleidomastoid
muscle between the mastoid process and the medial clavicle,
EMGsca was placed within the posterior triangle of the neck
at the level of the cricoid process, EMG7ic was placed in the
space of the seventh rib along the anterior axillary line, and
EMGpic was placed in the space of the second rib roughly
3 cm lateral to the sternum.

The pressure differences also supported the abdominal breathing causing better diaphragm usage:
 The catheter used for EMGdi
includes two balloons for measuring esophageal (Peso) and
gastric pressures (Pga) simultaneously. Each balloon was connected
to a differential pressure transducer (model DP15-34;
Validyne Engineering, Northridge, CA). Transdiaphragmatic
pressure (Pdi) was calculated as the difference between Pga
and Peso. Absolute pressure swings were calculated as the
difference between the maximum and the minimum pressures
during each inspiration. Pressure–time products (PTP) of Pm
(PTPm), Peso (PTPeso), Pga (PTPga), and Pdi (PTPdi) were
obtained by calculating the area subtended by each pressure
trace as a function of inspiratory time.

The rate of force development (pressure time product PTP) was also in a similar direction:

And finally, a single users data:
I highlighted the diaphragm pressure in yellow.  The IMTdi (right side) shows more EMG activity, diaphragmatic pressure than the standard IMT.

The discussion has several important points including:

Our findings suggest that a typical session of IMT requires
a high level of activation in the scalenes and sternocleidomastoid
to assist the diaphragm
. In fact, EMGsca and
EMGscm were virtually the same as EMGdi when expressed
as a percentage of maximal activation (Fig. 1) with lower
relative EMG activity in the seventh intercostal space and
parasternal intercostals. These observations raise the question
of whether or not IMT could be improved to more effectively
target the diaphragm
. There are several lines of
evidence suggesting that the diaphragm should be the most
important muscle to target during IMT.

For example, diaphragm fatigue
can cause a redistribution of blood flow from the locomotor
to the respiratory muscles (33), known as the
respiratory ‘‘steal’’ phenomenon.
This sympathetically mediated
metaboreflex response reduces oxygen delivery to the
working locomotor muscles leading to increased locomotor
muscle fatigue, increased effort perception, and ultimately
reductions in exercise performance (28). 

In summary, this study demonstrates that inspiratory muscle
recruitment during IMT can be altered significantly based
on the instructions given to participants.
A typical bout of
IMT relies on the extradiaphragmatic inspiratory muscles to
assist the diaphragm. This may not be the optimal form of IMT
given the established importance of the diaphragm during
Simple instructions to encourage active recruitment of
the diaphragm during IMT can substantially increase the pressure
production and electrical activity of the diaphragm. Future
studies are required to determine whether IMTdi confers greater
benefits than IMT on exercise performance, diaphragmatic fatigue,
and physiological and perceptual responses to exercise.

Focused training with biofeedback:

The following RMT sessions were done using the Powerbreathe K4 with a reduced pressure setting from what I usually work with.  It became clear that I have a long way to go to train myself to target the diaphragm since the standard resistance was way too high for me.  As in strength training, it seemed reasonable to get my "form" proper with light weights before embarking on more challenging loads. 

Using the Hexoskin shirt for diaphragmatic breathing feedback:
Now that we've established that it is possible to target the diaphragm by altering the way we breath, can we do a better job in teaching ourselves how to do this?  In addition, since we don't have 3 hands, is there a way to observe the motion of the chest and abdomen as we do the RMT device?  The answer is yes to both and it involves the Hexoskin shirt (see bottom of post for further links).

Since the shirt is able to calculate the total ventilation rate, it needs plethysmographic sensors in both chest and abdominal areas corresponding to where the hands would go in the above reviewed paper.  So by looking at the chest vs abdominal excursions in the smartphone app, we can view our breathing pattern in real time.

How to set this up: 
Make sure the screen time out is adjusted - up to 10 minutes or so.
Open the Hexoskin app.
Start your RMT app/device.
This is the main screen of the android app (iPhone should be the same)

Once you have the shirt on, the module connected, press the show sensors button (in red).

After doing this, a screen with tabs showing either "Heart" or "Breathing" will appear.  Choose "Breathing" 

I have circled in yellow a normal breath (both inspiration - line going up, expiration - line going down).  The Thoracic/chest tracing is on the top, the Abdominal is on the bottom.
This is a normal deep breath and the thoracic/chest has greater motion than the abdominal.

Here is the transition of RMT training from "standard" to "diaphragmatic".  The top line flattens since I'm not using the chest (much), but the lower line representing abdominal.

The idea to is to minimize the thoracic motion and maximize the abdominal motion.  Just as in the article's instructions, you need to breath in making your belly protrude.  Kind of the opposite as sucking in your chest to show off at the gym.  We do not want much chest motion.  Before using the RMT resistance device just try breathing this way for awhile and get used to it.  Once it becomes more familiar, then try some very light resistance RMT.  But realize, it may take weeks (or months) to truly succeed with this technique.

This pattern continues through the training session:

The end of training, with a final deep normal breath for comparison:

Confirmation with the Hexoskin web view:
The Hexoskin website has a much more detailed view and analysis of measurements.
Here is a segment form the Hexoskin web app showing the end of a session of RMT using the abdominal breathing technique (guided by the smartphone app live view):
The abdominal motion is in light blue, chest/thoracic in dark blue.  The red circle is a typical breath in the midst of diaphragm training with excellent abdominal excursion, little thoracic motion.  On the right side are two normal breaths, both lines moving in a similar fashion.  This is good confirmation, but not too valuable as a biofeedback aide.

Here is a video of a shortened RMT session, starting with normal breathing, abdominal focused training and then normal again.  It is a screen recording of my android phone running the Hexoskin app in "sensor display" mode.

  • The diaphragm is the primary muscle responsible for respiration.
  • There is a breathing technique that can target diaphragmatic function according to EMG and pressure dynamics.
  • It is also possible to use the Hexoskin shirt, with it's smartphone app as a real time biofeedback device for optimal abdominal excursion as well as minimizing thoracic/neck muscle usage.
  • So far, no long term study has evaluated if this technique will lead to superior RMT results.

Other related post about the Hexoskin:

Friday, November 16, 2018

Blood flow redistribution II - Monitoring non locomotor muscle O2 during exercise

Prelim Draft, to be updated

A main theme of this blog has been the over emphasis in monitoring locomotor muscle desaturation during lower body endurance sports for purposes of pacing, training and recovery.  In other words, using the saturation change in areas like the vastus lateralis or rectus femoris to create training zones or predict exercise recovery status.  My personal interest has been looking at the change in costal O2, where this muscle competes for cardiac output from the increased work of breathing (increase in muscle usage) especially at high intensity zones.  From my experience working with athletes using both the Moxy and Humon sensors, the costal location is technically difficult to assess with these units.  The BSX works quite well, but is not sold any longer.  
  • Are there any alternate sites that can provide similar data?  
  • If so, do they show similar saturation changes with intense exercise?  
  • Are the mechanisms similar?

In a previous post, I explored the use of the deltoid area, which seems to track very closely to the costal O2.  Recently I was made aware of the possible use of muscle O2 in elite runners and decided to look into the O2 saturation changes in the suspected location that they are using.  Here is a photo of a runner using some sort of device, possibly an O2 sat sensor:

My initial thought was that the band looked too thin to be an O2 sat device, but if a tiny startup company like BSX could develop a small sensor, imagine what Nike or that level of resource could do several years later.  As an aside, that was one of my disappointments with the Humon hex, way too big and clunky given how compact the BSX design was years before.  The Moxy is a different animal entirely, it looks like a home built prototype, but was revolutionary at it's launch (like the old "bag" style cell phones).
However, as it really turned out, the pictured device was a heart rate monitor.

  • So the issue remains, is this a viable area to monitor, and if so, what do the tracings look like? 
  • How do they compare to the costal patterns that some feel represent a reasonable way in demonstrating cardiac output redistribution during intense exercise?  
  • If these non locomotor muscles do desaturate, can we differentiate between blood flow redistribution away from non essential areas and local contractile effects?  This is an important issue, especially in cycling where just gripping the bars may create muscular external compression related desaturation and perhaps an increase in usage (as in strength training).
  • If there is significant shunting of flow away from arms, is that an optimal area to be placing an optical heart rate monitor even though that particular sensor is targeted for skin tissue (not muscle)?

During muscle exercise, the circulatory system routes blood flow toward the areas that have increased need.  Since the pump (heart) has a fixed maximum rate of flow, shunting away from various areas is needed to maintain a reasonable blood pressure.  For example, in a house with a fixed flow well water pump, turning on all the faucets and sprinklers will make it very difficult to take a shower with ample flow.
A recently published synthesis of this subject makes an excellent case in the importance of blood flow competition between respiratory and locomotor muscles in endurance sports and fatigue.
We need to go back a few years to find a review of the interaction of upper and lower extremity muscles during exercise.  From that paper:

During the review of systemic oxygen consumption, cardiac
output, and values for peak skeletal muscle blood flow, one
key discrepancy emerged. Values for skeletal muscle blood
flow in quadrupeds, including dogs and ponies that are
considered “athletic” animals, are higher than values for
skeletal muscle blood flow seen during large muscle mass or
whole body exercise in humans. However, they are similar
to the values seen in humans during one-leg knee extension
exercise. This suggests that blood flow to contracting human
muscles is restrained during large muscle mass or
whole body exercise. This is true even in elite human athletes
with very high cardiac outputs. As shown in FIGURE 8,
if blood flow to the arms and legs during whole body skiing
had been similar to the values seen during either arm or leg
only skiing, then mean arterial pressure would have fallen
to 75 mmHg versus the observed 95 mmHg assuming
no change in cardiac output. The observation that blood
pressure did not fall in the skiers can be explained by restraint
of blood flow to the contracting muscles under these
circumstances. This discussion provides one line of evidence
that there can be competition among systemic blood pressure
regulation, cardiac output, and the demand by contracting
skeletal muscles for blood flow during heavy large
muscle mass or whole body exercise in humans

No wonder skiers have such high extraction rates!  The active muscle feed vessels are themselves constricting (competing with multiple other areas of usage).  In contrast, during cycling, the legs can get a larger portion of cardiac output at the same overall effort (flow does not need to go to the arms) therefore the O2 extraction does not drop to the same levels.  At max skiing power, either a drop in BP (and resultant loss of brain/vital organ perfusion) will occur, or the feed vessels need to narrow.
In exercising and non active muscle there is continual regulation of flow, and by inference, the degree of O2 extraction.  As flow drops (to maintain a reasonable BP), the O2 extraction should rise, with the same degree of muscle oxygen usage:

Further from the paper:
D. Arm Exercise Added to Cycling Can
Reduce Leg Blood Flow
When arm exercise of sufficient intensity is superimposed
on ongoing leg exercise, there is some evidence that leg
blood flow declines so that cardiac output is “conserved”
when total blood flow demand by the exercising arms is
increased (422). Likewise, maneuvers that increase the
work of breathing and thus the demand for blood flow by
the respiratory muscles have been shown to cause reductions
in leg blood flow during heavy exercise (197). However,
as noted earlier, the effects of adding arm exercise to
ongoing leg exercise can be complex and contradictory.
This is likely the result of differences in the study protocols
and the extent to which the subjects had both highly trained
arms and legs (67, 413, 422).
So the reverse should also be true, cycling should reduce flow to the arm muscles (and elsewhere).

Finally, no matter what the potential demand is for cardiac output, first and foremost priority is brain perfusion (along with other vital organs), through maintenance of blood flow and pressure:

With respect to the normal physiology above, lets now look at some non exercising muscle behavior during various cycling intensities.
Given the way I record the data from 3 O2 sensors and cycling power, the graphing is less than optimal.  It consists of a combination of Garmin Connect and tracings.

(To see the anatomy, structure and function of some of these muscles look here.)

Let's start with a comparison of the deltoid with the rectus femoris and costal areas.
Although this has been looked at before, more data can always be helpful.
The tracing below is the same presented in the VO2 max post, a 3 min 350w interval, followed by 30 second intervals with short rests.
Costal O2 yellow, RF O2 blue, Power red
Although the exact degree of desaturation is different, there is general agreement in the up and down pattern of the deltoid, rectus femoris and costal O2 placements.
Deltoid O2 yellow, Power red

The next was presented in the Bohr effect post, a VO2 max effort, followed shortly by a lactate threshold interval.
Costal O2 yellow, RF O2 blue, Power red
Costal and RF:

O2 sat yellow
  • The deltoid closely resembles the costal curve.  
  • In addition, the values remain stable in an intermediate zone on the second interval.

Lastly, the same as above, but a longer rest between intervals:
Costal O2 yellow, RF O2 blue, Power red
Deltoid O2 yellow
  • The deltoid once again mimics the shape of the costal desaturation.
  • However, a mimic of the curve does not mean the blood flow/muscle activity dynamics are the same.

What about blood flow priority?
Since we are monitoring O2 extraction and not flow, the blood flow to the deltoid may have been controlled differently (restricted) than the the costal muscles and this effect is not noted by the tracings above.  The deltoid is presumably doing little muscular work, yet the extraction drops as quickly as the costal area which is working quite hard with breathing.  As other studies have shown, costal flow is rising despite the desaturation.  So both areas are desaturating but for somewhat different reasons.  Costal muscles are actively working (like the arm/leg diagram above) and getting more flow, the deltoid is probably just getting reduced flow from baseline.

If the deltoid is a lower priority area for perfusion than the important accessory respiratory muscles, one may see some difference in re oxygenation patterns after a max effort with continued pedaling.  The costal area could (should?) have a faster recovery than the non essential deltoid.

Costal vs Deltoid patterns after 1 min maximum plus continued effort:
What I am trying to induce here is a situation where the cardiovascular system is maximally taxed, so output is at highest possible.  Then after the circulatory system auto regulates flow accordingly, is there a difference in how quickly re oxygenation occurs in each site?
As stated before, we are not measuring flow.  Since the costal muscles are still working near max (near max ventilation rates), quicker reoxygenation here would imply continued high flow.  Continued high muscle usage without enhanced flow should present as a delay in O2 sat rise (as in heart failure or occlusion).  On the other hand, the arm muscles not directly involved presenting a tempting target for the circulatory system to vasoconstrict, and therefore continued hypoxia (even though they are relatively inactive).

The O2 saturation curve rises pre-interval during coasting, then drops with the maximal effort.  There is about a 1 minute low level soft pedaling phase then a 250w x 2 minute final interval, with some soft pedaling after.
Costal O2 yellow, Power red
VS the Deltoid pattern (same interval, just not on the same graph):

  • Both sites drop to low O2 saturation levels.
  • The costal O2 recovers to pre interval baseline by 1 min of intermittent pedaling (green line).
  • Deltoid O2 saturation does not approach pre interval baseline after 1 minute (lower green line).
  • After the second interval, costal O2 overshoots to the level seen with coasting (green circle).  This may represent post muscle exertion vasodilitation.
  • In comparison, after the second bout, the deltoid does not overshoot to the coasting saturation value (green circle vs upper green line).  Since it was not active, there is no reason for dilating these vessels.
  • It seems plausible that the deltoid is not getting the same flow as the costal during recovery, resulting in a delay in oxygenation values.

Inner vs outer forearm vs Biceps - issue of local usage:
The deltoid has been noted by myself and others to be a reasonable surrogate for monitoring cardiac output redistribution while biking.  Do other muscles in the arms perform the same way?  This issue is somewhat complicated by the fact that just holding the bike steady under extreme load may introduce upper extremity muscular contraction of a significant enough amount to desaturate the forearm muscles.  One way to look at that effect is by observing the total hemoglobin.  As was seen in strength training, with a high enough load, external muscular compression will sharply reduce the THb, which is then followed by a rebound rise.  Let's look at this possibility:

The following 2 tracings were done during a 1 minute max, all out interval (exactly like others presented in other posts) with a continuation of pedaling at a low power (about 180w). This presumably represents a near max cardiac output and high respiratory muscle usage situation.
The green line represents O2 saturation and red is the total Hb.  Power is the background gray.  The first figure was from a sensor on the biceps:

  • There is an excellent desaturation response in the biceps to below 5% that continues on into the easy pedaling phase.  
  • Although there is some decrease in total Hb, it is rather mild and then slowly rises into the recovery.
Is the severe hypoxia in the biceps related to external compressive effects?
Let's compare the magnitude of total Hb change in the cycling biceps session above with one of my old strength workout tracings.

The above 1 min max cycling session:
Red is THb, Green is O2 sat
  • Yes, there is a drop in total Hb with the initial high power start, but during the low power second part, the THb rises, yet the hypoxia continues. (blue arrows)
  • I would have expected some return of the O2 saturation during that low power interval if the initial hypoxia was purely related to local bicep muscle usage.

Strength training biceps curl with 50% 1 rep max load for 45 seconds x 3 sets:
  • The biceps O2 saturation and total Hb clearly drop sharply with each set
  • Even within each set, one can appreciate some reflow during each repetition as the muscular compression changes.
  • The net range for total Hb drop in the bicep curl is almost triple that of the cycling interval (13.0-12.0 vs 12.55-12.2).
  • So although the cycling biceps tracing appears to have some total Hb fluctuation, it is of a much smaller degree than it is capable of showing.
Therefore it appears that the biceps saturation tracing during maximal cycling is minimally affected by local muscular use/compression.

    The external forearm saturation and THb response is somewhat different.  Here there is a sharp decrease in both O2 saturation (in green) and THb (red line, blue arrows), very reminiscent to that of strength training.  The gray background is the costal O2 which does not drop as quickly but seems to recover in a similar pattern.  Remember, the drop is related to flow as well as usage.

    From the above it seems that the biceps is not as prone to muscular compressive force as the forearm.  However, even on very easy cycling (160w) there is some transient total Hb shifting in the biceps.  The gray background is power.  Note is made of a total Hb drop at each resumption of pedaling as well as a rebound when coasting but the maximum differential is only 12.5 to 12.3.  At this power and bar grip force, I don't believe local muscular dynamics are the reason. 
    Therefore, the question of local muscular compressive force playing a role in arm muscle desaturation is still a theoretical issue.  From the common sense viewpoint, at max power levels, bar grip and arm stabilization must be factors, but at low levels of cycling power, it is hard to reconcile.

    External Forearm O2 saturation dynamics:
    Here is a more complete look at external forearm vs costal saturation.  Similar to previous, a one minute max followed by an "easy" 1 minute of cycling.  The external forearm quickly drops to near zero and remains near zero even with the easy phase, whereas the costal drops during the max effort, but recovers even during continued submaximal pedaling.
    Costal yellow, External arm blue, Power red, THb purple

    In terms of cardiac output redistribution, one could make a case for a hierarchy of tissue perfusion timing.  The active muscles are competing heavily with the respiratory muscles (with respiratory perhaps the highest priority), so costal flow would easily take priority over the "non essential" forearm (or biceps).  That the Forearm re-saturation mirrored the costal, does not necessarily mean that the mechanism is the same (as explained above).

    Here is a look at the External forearm vs costal during a VO2 max 3 min (about 350w) followed closely by a lactate threshold interval (about 250w):
    Costal O2 yellow, External arm blue, Power red
    • Just as in the maximal interval, the external forearm O2 saturation drops rapidly and approaches zero (while costal was 68 to 17%), and both the costal and outer forearm were both gradually dropping in the second interval (250w).  
    • The O2 recovery was faster in the costal area after both sessions.

    Inner forearm O2 saturation dynamics:
    Low power observations-
    Looking at some baseline and low power tracings, the inner forearm does mildly desaturate with relative stability from 67 to 60% at a half VO2 max power: 
    Yellow is O2 sat, Red is power

    At higher power (just below lactate threshold), there is further mild desaturation
    66 to 56%:

    During a VO2 max then lactate threshold interval (350w then 250w)
    80 to 5%:
    • The inner forearm certainly desaturates well during the higher power effort.  
    • During the lactate power interval there is prominent but stable desaturation.

    Maximum effort effect on Inner forearm:
    Yellow is Forearm, Power red, THb purple 
    • Here we see a very steep drop in saturation (80 to 4%)
    • Brief transient drop in THb (?external compression)
    • Prolonged desaturation after the maximum effort (about 13 sec at 4%) then a persistent degree of hypoxia during easy pedaling.  I purposely had a very loose grip on the bars during this time.  This area is probably not getting good perfusion as opposed to the costal muscles.
    • The THb sharply rises during the easy interval, but the hypoxia remains significant.

    Here is another 1 minute max (different day), followed by a 2 min interval at 190w inner forearm data:
    • The forearm muscle hypoxia continues into the post max session (first blue arrow), despite total Hb rising (purple arrow).
    • Inner Forearm re-saturation starts in earnest only with coasting (2nd blue arrow) 
    • Total Hb shows some minor fluctuations but does not seem to make much impact on O2 sat.
    • As opposed to the above - The Costal O2 sat rose rapidly during the lower power interval (below):

    Biceps O2 saturation dynamics:
    Low power observations-
    The low power tracing looks comparable to both the forearm and costal (all about 8% desaturation): 
    Costal O2 yellow, Biceps O2 blue, Power red

    During higher power (just below lactate threshold):
    Both costal and biceps become slightly lower than the 160w zone and steady.

    A slightly different variation of interval power modulation with interesting possibilities:
    The first minute is at 400w (well above VO2 max) then 2 minutes between VO2 max power and lactate threshold.
    Costal O2 in yellow, Biceps O2 in blue, Power in red
    • Both sites desaturate during the initial high power phase, however in the more moderate power phase they diverge.
    • The costal pattern stabilizes when the power is backed off.  Although not shown, ventilation rate was near maximal.  Presumably both blood flow and usage are high.
    • The biceps continues to fall during the 280 watt phase and ends at a value of 18%.  Instead we may have a low usage, low blood flow condition.
    • Does this represent an example of a higher priority of flow to the costal area, maintaining saturation of legs/respiratory areas at the expense of the "non essential" biceps.
    • Can this be used in optimizing racing or zone training choices? 
    • Can variations on this principle be use to gauge recovery ability?

    Maximum 1 min pattern in the Biceps:
    Costal O2 in yellow, Biceps O2 in blue, Power in red

    • The costal O2 drops but recovers during the low power phase after a maximum 1 minute effort (probably from continued high flow).
    • The biceps O2 drops ending up near the zero range and stays there during the low power interval and does not recover until coasting
    • Another possible example of blood flow priority redistribution with the biceps very low on the list.  Especially evident in recovery.


      Tracing of biceps (different day) during a 1 min max, then 190w x 2 minutes:
      Biceps O2 in green, Biceps THb in red, Power in gray
      • Extreme hypoxia continues (1-3%) during the second continuous 190w interval and does not abate until pedaling stops.
      • The Total Hb rises during the low power interval making the biceps hypoxia unlikely to be due to muscular compression (see below).

      Practical aspects in the field:
      The observations above are fascinating from a basic science perspective but can we use the differential desaturations of various non leg muscles to gauge effort and recovery.  The latter is of great importance for both interval repetition training as well as how long to wait until hammering a breakaway attempt.  Waiting in the pack for an extra few minutes could be a strategy if the biceps or forearm are still at diminished saturation.  However, if the biceps/forearm are back to baseline it may be time to go.
      I decided to try to simulate a short breakaway by doing 1 minute at 380w (10% over VO2 max power), then about 7 minutes at 270 (10% over MLSS).  At this point the power was backed down to 180w to see where the various muscle O2 values would rise to.  Then after just a 2 min active rest, repeat 180w x 6 min to observe the values again.  
      Power in red

      My thoughts were that:
      • Costal O2 would fall during the intense first minute, then stabilize at 270w, then return to baseline at both the 180w sessions.  I have done this sort of pacing many times, and that is what generally happens.  The Costal O2 is helpful at high intensity, but less so at MLSS.
      • The forearm and biceps would fall during the intense 1 min but continue to drop (unlike costal) during the 270w continuation.  If this is so, having a tool to help modulate the power around the critical area of MLSS could be very helpful.
      • In the forearm and biceps, there would be some return in O2 saturation during the 180w, but not to baseline.
      • Differential response after the short rest - After the short 2 minute active rest, the next round of O2 desaturation values in the forearm/bicep would be milder than the pre-rest segment of the same power.  The reason - less need for blood flow diversion away from those sites, since a short rest took place allowing some return to homeostasis.
      Biceps/Costal patterns:
      Costal O2 in yellow, Biceps O2 in blue, Power in red
      • Both costal and biceps O2 saturation rose in the zero power coasting phase leading up to the initial fast start.
      • The costal O2 did dip then rise (top blue arrow) into a stable zone during the initial higher power phase.
      • During both the first and second 180w zone, costal O2 was near identical and at baseline (top black line).
      • The biceps O2 saturation dropped initially (from a high of 58%) and continued to fall (bottom blue arrow) throughout the high intensity zones to a minimal value of 8%.
      • During the first 180w interval (no rest), the biceps O2 slowly rose to 30%.
      • However, after a short rest the biceps O2 at the same 180w was much higher remaining steady at 38% (thin black lines).  A differential of 8%.
      • It would seem logical that there would be a more significant blood flow diversion away from the arms in the first 180w bout than during the second occurring after a rest.

      Inner forearm (same interval as above):
      Forearm O2 in yellow, Forearm THb in purple, Power in red 
      • The forearm O2 sat dropped rapidly with the first minute but surprisingly remained stable during the 7 minutes of moderate intensity.  The pattern is somewhat between the costal (rising) and biceps (decreasing).
      • The forearm O2 sat rose rapidly during the first 180w interval to 47% (lower blue line) but.....
      • As in the biceps, the second post rest identical power session allowed a return to a higher degree of saturation at 53% (top blue line).  
      • The lower O2 saturation at the same power level but without a rest does seem to be a consistent pattern in both biceps and forearm but not the costal muscles. 
      • It is possible that either Biceps or Forearm O2 saturation could be a useful tool for interval recovery.

      Final thoughts:
      • Monitoring of non locomotor muscle O2 saturation is feasible, and may provide insights into cardiac and circulatory physiologic status.
      • The desaturation/total Hb profiles of the inner, outer forearm, deltoid, costal and biceps exhibit some similarities as well as interesting differences.  Even though the tracing appearances may be similar, the mechanisms involved are different.  Changes in extraction can be caused by alteration in blood flow, muscle usage and muscular compression.
      • One would expect the respiratory accessory muscles to have a higher priority than non exercising muscles for blood flow.  Although despite high activity, they may not desaturate faster than arm muscles because of appropriate flow enhancement  They do appear to re saturate faster (again from high flow).
      • The decrease in flow to the non exercising muscles (in the arm for example) should result in O2 saturation dropping at this area from higher extraction.
      • Different recovery milestones may be observed depending on the area measured (costal, deltoid, inner forearm, legs).  The area lowest in priority should resaturate last, heralding return to homeostasis.
      • Flow distribution away from the arm at high exercise intensity (as well as motion) makes this site less than ideal for optical plethysmographic heart rate placement. No wonder OHR at the wrist tracks so poorly at high intensity but is fine at rest. Vendors should consider other sites (ears, forehead) with less compromise in flow, as well as motion artifact.
      • Real time usage by athletes during an event - observing muscle oxygen saturation of non locomotor areas could provide information about overall exercise effort and the body's recovery strategy . It may be possible to leverage the monitoring of these sites for optimal zone training, race pacing, recovery dynamics and improvement in cardiovascular status.