Update to ‘Prescribing VO2max’

My last post left off with a novel workout protocol for optimizing VO2max training, according to my synthesis of the literature and personal experimentation at the time. I was blown away by the amount of engagement and discussion generated by the post. The discussion helped me expand my own understanding, challenged my assumptions, and of course left me with even more questions to explore.

This Reddit thread in particular garnered a bunch of very interesting discussions and is very worth reading. And TrainerRoad even implemented a similar workout in their program!

TRR Rattlesnake
Gotta admit, it was really cool to see this workout pop up. Great minds think alike?

Let me address where my assumptions have changed. If you haven’t read the last post on Prescribing VO2max, this post might not make much sense, so go read that first.

Assumptions Reconsidered

I think the proposed HR target of 88% HRmax was too low for most people to achieve >90% VO2max. 90% HRmax seems to be a better target for estimating 90% VO2max, although even this target may be insufficient for some people. [UPDATE: See how well VO2 & HR correlate for a couple different athletes here in a real training environment.]

VO2-HR relationship seems to be too variable for precise prediction. It depends on the individual, their current training/fatigue/stress, and most importantly the specific workout protocol attempted.

I still think using a Hard-start protocol will allow greater accumulated time above >90% VO2max, despite potentially accumulating less work (kJ) at VO2max power. I’m satisfied by the theory, and with demonstrations by Billat et al, 2013Lisbôa et al, 2015, some private experimental data (see below) [UPDATE: and now some personal experimentation with VO2 measurement] that this is a valid assumption.

Lisbôa et al, 2015

I’m not certain as to whether continuous (eg. 4×5/2.5min) or intermittent (eg. 3×13 30/15s) intervals are better for maximizing time >90% VO2max. The results of the more recent Rønnestad studies (2013 & 2015) are very compelling, but the findings of one of the more definitive review papers (Buchheit & Laursen, 2013) was equivocal, concluding that: “the most efficient way of accumulating a given [time >90% VO2max] in a HIT session (ie. intermittently vs. continuously) is still to be determined.” (p 329). Once again it depends on individual variability in VO2 kinetics, fatigue, fitness, etc. Keep reading some further suggestions.

Some Revealing Raw Data

Raphael at Diagnose Berlin was kind enough to share some very compelling experimental data investigating a few different workout protocol. His findings suggested a few things:

First, observing VO2 & HR under various workout protocol with continuous vs intermittent, and Hard-start decreasing-power vs steady-workload intervals showed that there is no consistent relationship between VO2 and HR. Certain workouts produced a large volume of time above >90% HRmax, but barely any time >90% VO2max.  Therefore %HRmax cannot be used to accurately or consistently predict %VO2max, even for the same individual within the same workout.

Second, Hard-start intervals elicited 90% VO2max more rapidly than intervals at a steady workload, both for intermittent and continuous protocol. This makes sense, as increased intensity will cause faster VO2 kinetics (faster rise toward VO2max). The trade-off is that a constant higher intensity will also necessarily shorten time to exhaustion. So the Hard-start interval must find a balance between sustaining near-maximal intensity and VO2, without succumbing to exhaustion.

And finally, continuous intervals seemed to produce greater time >90% VO2max than intermittent intervals. In fact, the highest achieved time >90% VO2max for the athlete observed by the Diagnose Berlin experiments was with the continuous Hard-start workout protocol suggested by Lisbôa et al, 2015, while the Hard-start intermittent 30/15 protocol was the second lowest. These were the clearest data challenging the effectiveness of the 30/15 protocol.

Comparing Hard-start continuous intervals (top) to Hard-start intermittent intervals (bottom) shows a difference in both time >90% HRmax (559s vs 278s) and time >90% VO2max (364s vs 26s!)

I have to admit that seeing raw data for myself was probably more effective at challenging my assumptions than reading all the various conflicting interpretations in the literature. Anecdote was more convincing than evidence.. which isn’t great, but often true.

Conclusions from Buchheit & Laursen, 2013

Buchheit & Laursen, 2013 highlight possible interval prescription as suggested by current best-evidence. Although they point out the lack of evidence as more of a limitation. Let’s take a look at the conclusions from their review.

Relevant concluding quotes from Buchheit & Laursen, 2013

  • “Short work intervals… intensity [should] range between 100% and 120% of vVO2max” [velocity at VO2max in running, analogous to power at VO2max in cycling]. (p 325)
  • “In [less well-trained] cyclists (VO2max: ~52 ml/min/kg),… larger volumes of HIT performed at lower intensity (ie. 4x8min at 90% HRmax) may be more effective than more traditional HIT (eg. 4x4min). Further research… in more highly trained athletes are required”. (p 325)
  • “Longer work intervals are preferred for individuals with slower VO2 kinetics (ie. older/less-trained), or for exercise on a bike” 2013, (p 326)
  • “A direct comparison between long [continuous] and short [intermittent] HIT sessions, with respect to [time >90% VO2max] has only been reported twice in highly trained athletes… [time >90% VO2max] was ‘very largely’ lower during the 30/30s intervals.” [this in comparison to 2/2min]. (p 328)
  • “If we consider a goal [time >90% VO2max] of ~10min per session is appropriate to elicit important cardiopulmonary adaptations… athletes should expect to exercise for 30min using 30/30s (110%/50% vVO2max), since [time >90% VO2max]/total exercise time ratio is approximately 30%… it is possible for it to be broken down into 3x sets of 10-12min”. (p 327)
  • “We recommend long- [>1-2min] and short- [<45s] bout HIT with a work/relief ratio >1.” (p 332)
  • “Near-to-maximal and prolonged [continuous] work intervals currently appear to be the preferred HIT option (ie. >4min @ 90-95% v/PVO2max, with likely decreasing external load [Hard-start] with increasing fatigue to prolong [time >90% VO2max].” (p 332)

There are definitely some conflicting conclusions based on the literature available. The end result seems to be that there is no one best option. “most HIT formats, if properly manipulated, can enable athletes to reach VO2max” (p 331).

Current Thinking

I think the current evidence points toward continuous Hard-start decreasing-power intervals to optimally drive VO2 adaptations for most people. I certainly still think the intermittent Hard-start 30/15s will also be effective, however continuous intervals seem more compelling and better investigated in the existing literature.

The main limitation of prescribing VO2max intervals remains that HR and Power are not linearly related, nor well correlated to VO2. Power and HR targets derived from a single MAP or FTP test seem unreliable in predicting the actual VO2 response to a workout.

I need to continue to learn how VO2 kinetics responds to intensity and duration, under both continuous and intermittent workout protocol. If VO2 cannot be easily estimated or measured at home in training [UPDATE: Now it can!], we need to establish a test protocol that can reliably & accurately predict VO2 values via prescription of power & HR workout targets. Or possibly, look to other measurable metrics like SmO2 to establish training zones and interval targets?

I’m still reading, thinking, tinkering, and experimenting in the background.
I’ll try to keep the blog updated more often than every 6 months!


13 thoughts on “Update to ‘Prescribing VO2max’

  1. For central capacity maximizing time above 90% HRmax seems legit, But the training effect is not only for the heart and lungs. Much of the adaptation we seek is for the legs! So i still think maximizing the average power over your intervalls, no matter the length, should be a priority.

    Liked by 1 person

    1. Johannes, I completely agree that we should be aiming to maximize average power over the course of VO2max intervals, for every workout and regardless of protocol. I like Rønnestad’s approach of targeting %HRmax and observing avg power as a descriptive outcome measure: ie. as you get more fit, avg power will obviously go up for the same given HR.

      However I wouldn’t say targeting HR is only/primarily for central adaptations, and Power for peripheral. VO2 quantifies systemic O2 utilization and the limiting factor to VO2max tends to be O2 delivery (ie. red blood cell & haemoglobin volume). “training-induced increases in VO2max are largely facilitated by expansion of red blood cell volume and an associated improvement in stroke volume” (Lundby et al, 2017. https://www.ncbi.nlm.nih.gov/pubmed/27888580) which are more central.

      I’m not sure how to best drive adaptations to either peripheral or central. My guess is directionally you’re on the right track: that intensity may drive more peripheral changes, while sustained duration drives more central. But I haven’t encountered a firm conclusion in the literature. And I suspect the mechanism behind ‘improving the legs’ may be more from anaerobic/lactate metabolic adaptations.

      Take a look at MacInnis & Gabala (2017. https://www.ncbi.nlm.nih.gov/pubmed/27748956). They put together a super in-depth discussion of various mechanistic outcomes from HIIT vs moderate intensity protocol. My takeaways were: mitochondrial changes & VO2max in general respond better to intensity. Stroke volume, haematological changes, & capillarization may respond better to duration.

      Liked by 2 people

  2. Hi Jem, been loving the blog and love following you and your workouts. Incredible learning for me! I pulled the Lisboa et al, 2015 article and read it – it seems they write that DPO intervals allow the athlete to spend more time at VO2max than CPO, contrary to what you wrote. Am I reading this right? Their DPO protocol was 2 intervals – each interval consisting of the time it took to reach exhaustion at power needed to reach VO2Max. During the interval power decreased 4 times to 110& of critical power. The CPO (continuous protocol) was just 3 intervals using that same power for 60% of time it took to reach exhaustion at power needed to reach VO2Max. Interested in hearing your thoughts.


    1. Thanks Edmund. Oh I think I see the misunderstanding, because I’m 100% in agreement that decreasing-power intervals should allow greater duration >90% VO2max, and I’m pretty sure I wrote as such!

      I’m using ‘continuous’ to mean a solid (eg.) 5min interval, instead of intermittent (eg.) 30/15 on/off micro-intervals. And ‘steady-power’ or ‘steady-workload’ intervals to mean power remains constant for the duration of the interval.

      So to clarify, I agree that hard-start decreasing-power continuous intervals should allow the greatest duration accumulated >90% VO2max.

      And in fact, I just got my hands on a VO2 mask, and initial testing seems to show exactly that! eg. on a 3min interval, hard-start at ~430 W for 90sec and decreasing to 360 W for the final 90sec, VO2 ramped up to 90% VO2peak in <60sec, and remained above 90% for the full remaining duration (2min) despite decreasing power by ~15%! Very exciting experimentation going on. Expect some posts on it soon!

      Liked by 1 person

  3. Have you experimented and gotten data from VO2 max intervals done at max with decreasing quality throughout set? If HR and breathing remain very high (>90% VO2 max), then wouldn’t stress on system be great even though power is dropping?


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.