Interview with Cristian Osgnach on Metabolic Power
uac-books-banner

Interview with Cristian Osgnach on Metabolic Power

I had the pleasure to interview Cristian Osgnach, who is one of the leading researchers and pioneers of Metabolic Power concept, as well as involved in GPEXE project, which is 20Hz GPS performance monitor & analytical system for professional sports. I asked Cristian question ranging from what is Metabolic Power to the best GPS/Accelerometry metrics to collect in running-based team sports. Very insightful interview, especially for those involved in collecting and analyzing GPS/Accelerometry data. Enjoy!

Hi Mladen and hello to your many readers!

I tried to answer your questions hoping to avoid so much nonsense (a field where I’m a little champion, especially if I have to use the English language!). I know that “metabolic power” is a hot topic, recently it has attracted research interest and has also been integrated in many tracking systems. But I’m surprised that “ …when a new approach comes along that is an improvement on the established way of doing things (i.e. assess the workload sustained by the athlete), rather than looking at the difference or improvement between ‘existing model’ and ‘new approach’, they instead focus on the difference between‘new approach ’and‘perfection’!” (a friend of mine commenting a critical article about the metabolic power model).

We believe that the metabolic power approach could represent an interesting way to better understand the effort of the players in team sports. Of course, the model needs further improvements (some of which have already been introduced in our upgrades) and it cannot provide a feedback on ‘everything’… however, if the goal is to get a clear picture about the metabolic characteristics of training, I’m sure it can be much more useful than traditional methods based on speed and acceleration thresholds and/or categories.

Furthermore, it’s interesting to note that the majority of recent criticisms to the metabolic power metrics are exactly the same that could be made to traditional criteria based on speed or acceleration. For example, the estimated energy cost from Minetti’s polynomial, as described in our original paper of 2010, was calculated assuming a fixed terrain constant (KT = 1.29). One can discuss for a long time about the way to improve this approximation because KT may change a lot if we consider a compact terrain (i.e. lower KT → lower EC → lower MP → lower energy expenditure) or a muddy pitch (i.e. higher KT → … → higher energy expenditure). However, obviously enough, it can be expected that also high speeds and high accelerations are energetically more expensive on a muddy pitch, even if the analyses are based on fixed speeds and accelerations. Why in this case there is not any problem? Is it because the constant KT does not appear?

I hope that your availability to publish the following Q&A on your blog will really help the readers of your popular website to get a better idea on opportunities and limits of our proposal.

Thank you so much for the opportunity!

Metabolic Power (MP) is getting more attention lately. Can you please explain to the readers what is MP and how it is estimated? Is it, and should it be, related to VO2 consumption to be a valid metric? Is it better than velocity and acceleration metrics?

The best way to explain the meaning of MP (especially if we refer to team sports) is to provide a clear definition; therefore, MP is a measure of the overall amount of energy required, per unit of time, to reconstitute the ATP utilized for work performance. In running, as well as in any other form of locomotion, the product of the instantaneous velocity (v) and the corresponding energy cost (EC) per unit body mass and distance yields the instantaneous MP necessary to run at the speed in question:

MP = v ⋅ EC

Since many years, tracking technologies have allowed to calculate the MP thanks to the measurement of v and the estimation of EC. During constant speed running both v and EC are constant and MP increases as a linear function of the speed (see figure 1, red line); but this is not the case when either the acceleration (and hence the corresponding EC) and/or the speed varies, which is the typical scenario in team sports. In this context, accelerated/decelerated running on a flat terrain can be considered biomechanically equivalent to uphill/downhill running at constant speed, the slope being dictated by the forward acceleration. Since EC of uphill/downhill running at constant speed is well know, this made it possible to estimate EC of accelerated/decelerated running and hence MP from the measured acceleration (for further details refer to di Prampero et al., J Exp Biol, 2005). This is the way to estimate the MP in team sports.

 

fig-1

figure 1: metabolic power (Ė) and corresponding oxygen consumption (VO2) above resting as a function of the speed during three types of locomotion (c, constant running – red line; m, walking; m*, competitive walking).

 

During light intensity exercises, at steady state all the energy necessary for the development of mechanical power (i.e. energy demand) derives from aerobic sources: if this is the case, a simple measure of oxygen consumption (VO2) provides the overall energy needed for the exercise (i.e. energy supply). However, when the same concept is applied to team sports, very often the aerobic system is not able to provide all the energy needed for the exercise, either because the time is too short to attain the steady state, and/or because the exercise intensity is greater than the subject’s maximal O2 consumption; hence, the energy supply will be ensured thanks to a substantial amount of anaerobic energy. For this reason, it’s very important to use MP as a measure of the overall energy demand and VO2 only as a measure of aerobic energy supply: obviously, these two parameters cannot be used interchangeably if this is the context.

Is MP better than speed or acceleration? I don’t think this is the problem: each metric tells only what it can say! If the interest is for the high-speed, for sure speed is the parameter necessary for the analysis; but if the aim of the analysis is to detect the high-energy demand phases, definitely the MP will be the most appropriate. Often the confusion arises when the goal is to describe the ‘high intensity’: a general concept with no units, overambitious to describe with a single metric.

Why is polynomial MP better than just multiplying instantaneous velocity and acceleration(va)?

The product of speed and acceleration is the mechanical power. For sure you can consider this parameter to get some insights but you have to know the limits of this product:

1. if you move at constant speed, the mechanical power so calculated will be zero… but we know that the ‘effort’ at constant speed is not zero! By MP approach we avoid this problem because even if the athlete move at a constant speed we can get an estimated energy expenditure (see figure 1)

2. if you use (va), the acceleration activity have the same specific size that the deceleration activity (with the opposite sign); for this reason the MP polynomial take into account the extra-cost of acceleration while we know that the deceleration is not energetically expensive (in fact, we have to consider the ‘metabolic power approach’ as a metabolic/energetic analysis). On the other side, the contrary happens for muscular load because decelerations are more damaging on muscle than accelerations (therefore, ‘muscular power approach’ – a new model on which we have worked recently – is a mechanical/muscular analysis). Also in this latter case, when we use the muscular power, we have the opportunity to fix the zero load when the athlete move at constant speed.

With the recent development of cheaper and better accelerometers, do you believe GPS metrics (position derived, including velocity) will be a thing of past? For example, Catapult’s PlayerLoad and PlayerLoad 2D metrics are good proxies to physical load and are based solely on accelerometer data. Besides, the units might be way more cheaper and lighter.

Again, it depends on the purpose of the analysis. My hunch is that position (for tactical purposes) and speed and derived parameters (for physical purposes) will still be necessary in the future to obtain a complete analysis by the use of any tracking system (GPS, video, LPS, etc.). Accelerometers, but together with magnetometers and gyroscopes, definitely represent the way to improve accuracy of the wearable tracking systems which, as far as I can see, cannot be phased out so easily in the next future.

Furthermore, the interest in describing each single effort that contributes to the overall physical load (e.g. high speed events, high MP events, etc.) requires the availability of the time course of each single metric with their own units. All this information represents the core for a deep analysis and cannot be replaced by some generic metrics, expressed via arbitrary units, providing an overall load.

What is your opinion on using fixed vs. individualized velocity and acceleration zones? Can you tell us more about the novel “Percentage Acceleration” approach?

Fixed velocity and acceleration thresholds were the reason why in 2008 I tried to apply the metabolic power approach in soccer with professor di Prampero and Stefano Poser!

Focusing first on velocity, it’s not so easy to find a meaningful fixed threshold because there are no convincing (physiological) reasons that support this choice. For example, in many papers that deal with soccer, the fixed speed threshold to detect the activity called ‘sprint’ is placed at 25.2 km⋅h-1. But what happens when the speed exceeds this threshold? The player ran faster than 7 m⋅s-1 (an integer that is easy to remember and this is probably the reason for this choice) and… nothing else! Even when the speed threshold is individualized in some way, considering for example the maximal aerobic speed (MAS), it’s quite untrue that below MAS the activity carried out by the player is entirely aerobic. Similar remarks may be made about accelerations and this questions the use of speed and acceleration zones to get information about the overall concept of ‘high intensity’.

The metabolic power approach represents an attempt to overcome the limitations mentioned above because it is based on the following fundamentals:

  1. the threshold is an individual parameter represented by the maximal oxygen consumption (VO2max) of each athlete which is quite easy to measure directly or indirectly; it is also easy to measure it several times during the season;
  2. at any given time during the match or the training session, where the intensity change randomly, the energy demand represented by the instantaneous MP could be smaller, equal or greater than the VO2max;
  3. when MP exceeds the individual VO2max, the athlete needs an amount of anaerobic energy to sustain the intensity of the exercise;
  4. this amount of anaerobic energy can be defined as an oxygen debt which, at some point in the following phases of the exercise, must be paid;
  5. this is the reason why it make sense to consider all these occurrences as ‘high intensity events’ if the aim of the analysis is mainly targeted to the energetic balance of the drill in question.

Sonderegger K et al. in a recent paper in PlosOne proposed an original approach based on “percentage acceleration”. It’s not 100% clear to me how is possible to finalize the information gathered from this analysis in a general summary because it’s not so easy to define the ‘initial speed’ during the game. Nevertheless, I love the figure 1 of this article (reproduced for convenience below): in the lower panel it’s pretty clear how an acceleration of 3 m⋅s-2 (the threshold often used by many authors to detect the high accelerations in soccer, red dashed line) represent the maximum effort only when considering the greater initial velocity (i.e. 16.7 km⋅h-1); whenever the speed decreases, the same acceleration value (3 m⋅s-2) becomes only a fraction of the maximum force available to the player.

fig-2

figure 2: time course of speed (upper panel) and acceleration (lower panel); the acceleration peaks are strictly related to the initial speed from which the athlete start to accelerate maximally (reproduced from Sonderegger K et al. 2016).

There is some work on using accelerometer data to assess differences between left/right leg in straight-line running. Do you believe this have some practical merit in return-to-play protocols and injury prevention?

Not in the loop enough to know! My first impression is that the left/right balance could provide reliable information only under controlled conditions (e.g. in the lab, on a treadmill, on a perfectly flat terrain, during straight line run, etc.). On the contrary, this metric is usually collected during daily training sessions where the context is completely unstandardized. Just a couple of examples:

  • most of times the player stop-carry-kick the ball with the dominant limb
  • if the position of the player is on the left side, probably the running style will be altered because his body is often oriented toward the center of the pitch where the game is played
  • etc

What about the left/right balance in these circumstances? Will any step imbalance be due to variables related to the injury risk or it just depend on the specific context (e.g. dominant limb, position on the pitch, etc.)?

What would be the use of real time GPS analysis during a game? How far are we from estimating fatigue from these metrics during live performance? Is there any merit of these in judging player effectiveness/efficiency for the coaches besides tactical role and technical performance?

This is a very intriguing point! Personally, I have never used the real-time in the past… but there are some aspects that arouse my curiosity and I proposed to implement them in the GPEXE-LIVE system. Surely the fitness coach needs two basic prerequisites for the use of the live during the training: a coach open to this innovation/support and a system that shows consistent results when compared with the downloaded data at the end of the session (one of the main trouble of existing systems).

Once the above conditions are met, it could be very interesting to get an energy balance about the drill in progress. By monitoring together the time course of MP and VO2, it may be possible to estimate the emptying rate of the ATP+CP stores (i.e. related to the red area in figure 3, where MP > VO2 and an oxygen debt is contracted) or the refilling rate of ATP+CP stores (i.e. related to the green area in figure 3, where VO2 > MP and a substantial repayment of the oxygen debt occurs). This analysis, with some additional complexity that is not the case to itemize here, could lead to some useful information in real-time, especially referred to:

  • the involvement of the anaerobic lactic system which may compromise the capacity to maintain high intensity for long periods;
  • the glycogen depletion which also may lead to a reduction of the intensity;
  • the capacity of the player to perform as close as possible to his critical power.
fig-3

figure 3: time course of MP (red line) and VO2 (blu line).

Using the real-time as the way to encourage the laziest players to increase the intensity… otherwise, it seems to me a piece of nonsense!

What would be the few variables/metrics to collect in running-based team sports? What give the biggest bang for the buck in your opinion?

The amount of data collected with tracking systems is enormous and the challenge is to focus on few variables useful to monitor the workload. The analysis is not a self-seeking exercise but it must be an important addition to manage better the training (in other words, the best work/recovery balance to maintain or improve performance and avoid injuries). In order to select the winning variables, everyone looks for those that best represent the training load and… everyone have pretty good reasons to justify his own selection.

My personal scheme is to focus on the following points for each training session:

step info parameters
#1 quality of recovery questionnaire
#2 metabolic load metabolic power metrics
#3 muscular load muscular load metrics
#4 perceived exertion questionnaire

While parameters from step #1 and #4 come from elsewhere, metabolic and muscular metrics can be estimated via tracking systems with the aim to quantify the volume, the intensity, the characteristics of the training and events’ details (see the table below):

#2 metabolic metrics #3 muscular metrics
VOLUME: energy expenditure muscle load
INTENSITY: metabolic power muscle power
CHARACTERISTIC: EDI
(equivalent distance index)
EI
(eccentric index)
EVENTS DETAILS – events number
– work avg duration
– work avg power
– recovery avg duration
– recovery avg power
– events number
– work avg duration
– work avg power
– recovery avg duration
– recovery avg power

It’s not so easy to describe in few sentences the exact sense of all these parameters; metabolic metrics were discussed at length in our papers to which the reader is referred for further details (di Prampero et al., J Exp Biol, 2005; Osgnach et al., Med Sci Sports Exerc, 2010; di Prampero et al., EJAP, 2015); muscular metrics are a recent proposal and I hope that soon we will be able to publish something about them. Anyway, the overall goal of collecting these parameters is the possibility, thanks to time period summaries (from the single drill report to the overall week report), to monitor and manage better the workload for each athlete. For example, some practitioners/researchers do not consider useful (or possible) to plan the training by the use of metabolic power metrics. Is it really possible to work on the basis of Joules or Calories instead of time and distance? There are a lot of fitness coaches who build their workout plans using the total energy instead of the total distance as a measure of volume. The most daring of them even do the same for every single drill: it is just to be open to novelties, know the limits of the approach and simply benefit from the advantages. But if someone is frightened by this option… never give up time, distance and average speed!

You are involved in GPEXE product. Can you tell us more about it and why is it better than the competitor?

We have devoted a great deal of attention to the data acquisition characteristics, particularly to appropriately high sampling frequencies (20 Hz) and correct signal filtering. This allowed us to check that the resulting instantaneous speed turns out essentially equal to that obtained by conventional “gold standards” such as radar system for velocity control. In addition, we have checked that the estimated rate of O2 consumption during intermittent short terms high intensity exercise is essentially equal to that obtained by direct VO2 assessment by means of portable metabolic carts (di Prampero et al., EJAP, 2015). However, it should also be pointed out that this approach does not take into considerations several specific characteristics of soccer, such as jumps, running with the ball, kicking the ball or contrasting an opponent. Nevertheless we think that, in terms of the energetics of a match, it yields a picture closer to the “truth” than that emerging from more conventional approaches.

Thanks for insightful answers Cristian. Where can readers find more about you?

I like to occasionally share some thoughts on Twitter, although most of the time is really difficult to convey a clear message within 140 characters. But it’s got me thinking that maybe the space limitation is not the problem since after several years there is still a serious lack of clarity about the metabolic power approach applied to team sports! (laughing…)

 

I am a physical preparation coach from Belgrade, Serbia, grow up in Pula, Croatia (which I consider my home town). I was involved in physical preparation of professional, amateur and recreational athletes of various ages in sports such as basketball, soccer, volleyball, martial arts and tennis. Read More »
free-memeber-button

Home Forums Interview with Cristian Osgnach on Metabolic Power

This topic contains 0 replies, has 1 voice, and was last updated by  Mladen Jovanovic 3 months ago.

You must be logged in to reply to this topic.

Are you ready? Get it NOW!

Increase more than 700% of Email Subscribers!

Your Information will never be shared with any third party