Latest revision as of 11:08, 3 May 2020

You can think of Critical Power as a threshold between aerobic and anaerobic exercise. While that's an oversimplification, it reflects the subjective feelings of the athlete well. In some ways it's like Lactate Threshold, but easier to determine. Critical Power has been around since 1965^[1], but the rise of power-based cycling trainers, cycling power meters, and now the ability to estimate running power with Stryd, the concept has gained new traction. Critical Power is the highest workload you can sustain without using your anaerobic reserve, and research papers often consider this workload that an athlete can sustain "indefinitely." Or course, "indefinitely" here means until glycogen reserves get depleted, or other limiting factors set in. Typically, you can maintain Critical Power for 30-60 minutes ^[2]. Above the Critical Power, there's an amount of work you can perform anaerobically. This anaerobic work capacity is named W' (W-prime). You can either use your W' at high intensity for a short period or a lower intensity for a longer period. This creates a curve of how long you can sustain a power level above Critical Power.

Critical power, from Golden Cheetah.

1 An Analogy for Critical Power

Here's a financial analogy for Critical Power. You could think of Critical Power as your regular income, and W' as your savings. You can spend your regular income almost indefinitely and not run out of money. If you spend your money faster than you earn it, you use up your savings. You could spend your savings slowly for a longer period, or quickly for a short period. Once your savings are gone, you must spend less than your income to build your savings back up.

2 Why Should You Care?

There are several benefits to knowing your Critical Power as a runner.

• Improvements in your Critical Power show improvements in your fitness. This is a good way of evaluating the effectiveness of your training.
• Critical Power can help define your running paces. To do this, you'll need to find your Running Critical Power (see below.)
• You can use Critical Power for HIIT on a bike trainer. You can structure a HIIT workout based on your CP and W'. You can use these values to understand how stressful (and how practical) a workout will be. I use Golden Cheetah which will show the W' balance as a curve for a workout.
• The CP and W' can allow you to pace shorter races (30-60 minutes). An ideal pace will be one in which you deplete W' evenly over the race, reaching near zero at the end. For longer races, other factors beyond W' depletion become the limiting factors, so it's less valuable there. Again, you'll need to know your Running Critical Power.

3 Finding Your Critical Power

The gold standard for measuring CP is to perform multiple tests to define points in the CP/W' curve. You need to perform each test needs while well recovered, so each needs to be on a separate day. After a warmup, you do a fixed power workout that should cause your exhaustion between 3 and 15 minutes (durations less than 3 minutes may give an overestimate of CP^[3].) A total of five tests should be performed. Given these results, several online calculators will give CP and W' numbers, though I'd recommend https://shiny.fmattioni.me/CPapp/ which has multiple models, including the 3-parameter model if you have 4 or 5 test results. (One study found that CP occurred at around 83% of Max HR ^[4]).

3.1 The 3AOT Approach

There's a simpler "3 minute all out test" (3AOT) that can estimate your CP. This test must be "un-paced" where you go all out for the duration of the test. This means that power output will be very high for the first few seconds, then drop off as you deplete your W'. Between 90 and 120 seconds, you should have depleted all your W' and you can only maintain your CP. The limitation of the test is that any pacing, even subconsciously, will overestimate CP and underestimate W'^[5]. There are several studies that have found the 3AOT gives a good estimate of CP/W'^[6][7][8]. These studies found that a 3AOT can detect relatively small changes in CP from training, such as increasing CP from 230 to 255w^[9]. One validation technique that a study used was to exercise at 15w above and then 15w below the critical power estimate. The study found that athletes could not maintain 15w above their CP for over 10-15 minutes, where 15w below their CP was stable up to 30 minutes. However, one study found the 3AOT did not provide valid results in elite cyclists^[10]. That study found that the 3AOT estimated an average of 351w compared with 402w with multiple tests. This study found that the underestimated the athletes' W', measuring an average of 15.5 kJ rather than 24.3 Kj.

4 Problems with Critical Power

There are three problems with Critical Power: the simplifications of the model, the reality of an anaerobic threshold, and repeatability. Let's look at each in turn.

4.1 All Models Are Wrong

The quote "all models are wrong, but some models are useful" is appropriate here. There have been several extensions of the simple two parameter model^[11]. These extensions add more accuracy at the cost of complexity. Morton's 1996 proposal for a "3-parameter" model improves the estimates of CP and W'^[12]. There are online calculators that will use the more sophisticated models to give better estimates.

4.2 Anerobic Threshold

The idea of Critical Power, like Lactate Threshold is to find the cut off between aerobic and aerobic plus anaerobic energy production. In reality, there is no fixed line between the two. The definition of Critical Power is often in terms of the power output that can be kept up "indefinitely" ^[3]. But it's easy to see that pace or power output declines with increasing time in a curve with no sudden line that demarks an anerobic threshold. This means that the definition of an anerobic threshold tends to depend on how long "indefinitely" is. It seems that in the areas of anerobic thresholds, it's typically 20-30 minutes. There's some evidence that Critical Power corresponds to a 30+ minute endurance^[13].

4.3 Repeatability of Critical Power

A study of how training changes CP/W' performed a repeatability test^[14]. Twice they performed five bouts with exhaustion times between 1 and 10 minutes. Each bout was on a separate day in random order, and the five bouts were repeated a week later. They found that CP had good repeatability, with a correlation of the two tests of 0.92. The repeatability of W' was lower at 0.62. This suggests that tests using this approach to determine CP are valid, but caution should be used with measurements of W'.

5 Finding Running Critical Power

Most work on Critical Power uses a stationary bike. I found one study that used multiple runs to exhaustion^[15]. The study used 6 treadmill runs to failure in 2-12 minutes, each one separated by at least 48 hours. The treadmill used a safety harness, which reflects the problem with treadmill running to failure. It might be possible to do similar fixed pace outdoor runs using a vehicle for pacing. Unfortunately, that's not something many of us have access to. Another approach would be to self-pace using Stryd for accurate feedback but without a heads-up display, it would be tricky to maintain a steady enough pace for the results to be valid. Performing a 3AOT while running outdoors should be possible, but the all-out sprint at the beginning has high injury risk. Stryd will estimate your Critical Power, as described at [1], but I've not seen any independent validation.

6 Critical Power and Running Performance

A study looked at the relationship between Critical Power and Running Performance^[16] over 40m, 1km, 10km, and 21.1km (half marathon distance.) The study found a reasonable correlation between CP and both running performance and V̇O₂max for distances of 1km and longer. They found that CP correlated with 75-85% of both values (r=0.75-0.85).

7 Critical Power and Power at V̇O2max

I find it interesting that Critical Power is halfway between Lactate Threshold and the maximum power attained during an incremental V̇O₂max test^[17]. This suggests that the power achieved at V̇O₂max is partly from anaerobic sources.

8 Critical Power and Functional Threshold Power

A common power threshold in cycling is "Functional Threshold Power" or FTP. I recommend reading http://www.georgeron.com/2020/04/functional-threshold-power-FTP-myths-facts.html for a good analysis of the (lack of) scientific support for FTP.

9 References

1. ↑ H. Monod, J. Scherrer, THE WORK CAPACITY OF A SYNERGIC MUSCULAR GROUP, Ergonomics, volume 8, issue 3, 1965, pages 329–338, ISSN 0014-0139, doi 10.1080/00140136508930810
2. ↑ David W. Hill, The Critical Power Concept, Sports Medicine, volume 16, issue 4, 1993, pages 237–254, ISSN 0112-1642, doi 10.2165/00007256-199316040-00003
3. ↑ ^{3.0 3.1} D. Bishop, D. Jenkins, A. Howard, The Critical Power Function is Dependent on the Duration of the Predictive Exercise Tests Chosen, International Journal of Sports Medicine, volume 19, issue 02, 2007, pages 125–129, ISSN 0172-4622, doi 10.1055/s-2007-971894
4. ↑ Michelle Mielke, Terry J Housh, C Russell Hendrix, Jorge Zuniga, Clayton L Camic, Richard J Schmidt, Glen O Johnson, A Test for Determining Critical Heart Rate Using the Critical Power Model, Journal of Strength and Conditioning Research, volume 25, issue 2, 2011, pages 504–510, ISSN 1064-8011, doi 10.1519/JSC.0b013e3181b62c43
5. ↑ Mandell, Paul K., "Effects of a race timer on the three minute all-out test for critical power" (2019). Theses and projects. 261. https://digitalcommons.humboldt.edu/etd/261
6. ↑ Mark Burnley, Jonathan H. Doust, Anni Vanhatalo, A 3-min All-Out Test to Determine Peak Oxygen Uptake and the Maximal Steady State, Medicine & Science in Sports & Exercise, volume 38, issue 11, 2006, pages 1995–2003, ISSN 0195-9131, doi 10.1249/01.mss.0000232024.06114.a6
7. ↑ J. Dekerle, A. Vanhatalo, M. Burnley, Determination of critical power from a single test, Science & Sports, volume 23, issue 5, 2008, pages 231–238, ISSN 07651597, doi 10.1016/j.scispo.2007.06.015
8. ↑ Anni Vanhatalo, Jonathan H. Doust, Mark Burnley, Determination of Critical Power Using a 3-min All-out Cycling Test, Medicine & Science in Sports & Exercise, volume 39, issue 3, 2007, pages 548–555, ISSN 0195-9131, doi 10.1249/mss.0b013e31802dd3e6
9. ↑ Anni Vanhatalo, Jonathan H. Doust, Mark Burnley, A 3-min All-out Cycling Test Is Sensitive to a Change in Critical Power, Medicine & Science in Sports & Exercise, volume 40, issue 9, 2008, pages 1693–1699, ISSN 0195-9131, doi 10.1249/MSS.0b013e318177871a
10. ↑ Jason C. Bartram, Dominic Thewlis, David T. Martin, Kevin I. Norton, Predicting Critical Power in Elite Cyclists: Questioning the Validity of the 3-Minute All-Out Test, International Journal of Sports Physiology and Performance, volume 12, issue 6, 2017, pages 783–787, ISSN 1555-0265, doi 10.1123/ijspp.2016-0376
11. ↑ R. Hugh Morton, The critical power and related whole-body bioenergetic models, European Journal of Applied Physiology, volume 96, issue 4, 2005, pages 339–354, ISSN 1439-6319, doi 10.1007/s00421-005-0088-2
12. ↑ R. Hugh Morton, A 3-parameter critical power model, Ergonomics, volume 39, issue 4, 1996, pages 611–619, ISSN 0014-0139, doi 10.1080/00140139608964484
13. ↑ Dona J. Housh, Terry J. Housh, Sonja M. Bauge, The accuracy of the critical power test for predicting time to exhaustion during cycle ergometry, Ergonomics, volume 32, issue 8, 1989, pages 997–1004, ISSN 0014-0139, doi 10.1080/00140138908966860
14. ↑ G. Gaesser, L. Wilson, Effects of Continuous and Interval Training on the Parameters of the Power-Endurance Time Relationship for High-Intensity Exercise, International Journal of Sports Medicine, volume 09, issue 06, 2008, pages 417–421, ISSN 0172-4622, doi 10.1055/s-2007-1025043
15. ↑ R. Hughson, C. Orok, L. Staudt, A High Velocity Treadmill Running Test to Assess Endurance Running Potential*, International Journal of Sports Medicine, volume 05, issue 01, 2008, pages 23–25, ISSN 0172-4622, doi 10.1055/s-2008-1025875
16. ↑ Tracy Kolbe, Steven C. Dennis, Elizabeth Selley, Timothy D. Noakes, Michael I. Lambert, The relationship between critical power and running performance, Journal of Sports Sciences, volume 13, issue 3, 2008, pages 265–269, ISSN 0264-0414, doi 10.1080/02640419508732236
17. ↑ David C. Poole, Mark Burnley, Anni Vanhatalo, Harry B. Rossiter, Andrew M. Jones, Critical Power, Medicine & Science in Sports & Exercise, volume 48, issue 11, 2016, pages 2320–2334, ISSN 0195-9131, doi 10.1249/MSS.0000000000000939