Prescribing VO2max

Prescribing VO2max

June 26, 2018

I’ve written an UPDATE to this post where I reconsider a few assumptions given below. One major change is that 90% HRmax seems to be a better target than 88% HRmax. This post has been updated to reflect the changes.

Optimizing VO2max Interval Prescription

I left my conceptual post on VO2max with a tease for some practical application of VO2max intervals. Here’s where I’ll discuss some of the workouts I’ve been doing and the research that led me to them.

First, recall my definitions for T90VO2max and MAP (Max Aerobic Power).

T90VO2max is used to refer to the duration or time (T) spent at or above 90% VO2max, ie. duration at near-maximal rate of Oxygen consumption as measured in ml/kg/min via lab analysis.

MAP is maximal aerobic power, which I use to refer to the maximum power/workload, and therefore shortest duration that will elicit VO2max before ANaerobic energy system becomes the limiting factor to fatigue (read the last post for details).

VO2 Severe Moderate3_INKED2

Recall also that it appears that adaptations to VO2max are elicited by maximizing time spent above >90% VO2max (R√łnnestad & Hansen, 2013; Seiler, 2010; Thevenet et al, 2007; I could go on…) hence why I think T90VO2max is a valuable parameter of Aerobic/VO2max adaptations.

This is the value I’m trying to maximize for my VO2max workouts.

Thevenet et al, 2007

Intermittent Short Intervals

It has been shown that intermittent exercise can elicit greater total exercise duration at high intensities (R√łnnestad & Hansen, 2013;¬†R√łnnestad & Hansen, 2015;¬†Thevenet et al, 2007).

Conceptually this makes sense: having brief rest intervals should allow you to maintain the target workload for a longer cumulative duration than continuous exercise. This is basically just restating the concept of interval training.

The 2013 R√łnnestad & Hansen study referred to above found that 30sec work intervals with 2:1 work:rest ratio (30/15s) seemed to be most effective ‘short interval duration’ to maintaining longer T90VO2max durations.

Ronnestad 30s v tmax_INKED
R√łnnestad & Hansen, 2013

I wonder if 30sec as opposed to more individualized (percentage of Tmax), but less ’round’ numbers allowed the athlete to better pace their efforts over a known common duration?

They followed up this study in 2015 by comparing this 30/15 protocol with a standard 4x5min 2:1 work:rest protocol. They found the 30/15 protocol was more effective at improving VO2max and performance metrics after 10 weeks training.

Ronnestad SI v LI_INKED2
R√łnnestad & Hansen, 2015

Training with the¬†R√łnnestad 30/15s Protocol

The exact protocol R√łnnestad & Hansen used was 3 sets of 13×30/15s, with 3min RI between sets. They prescribed a workload target of >88% HRmax¬†(update: I now recommend >90% HRmax)¬†as a corollary to 90% VO2max. The instruction to subjects was to “perform at their maximal sustainable work intensity, aiming to achieve the highest possible average power output”

Power was used only descriptively as an indicator of performance, rather than as a prescriptive target: therefore if average power was higher for workouts later in the training program, this would indicate that the subject had improved their performance.

One of my recent R√łnnestad 30/15 workouts

I’ve used a modified R√łnnestad 30/15 Protocol on and off for the past year. It’s always a super tough workout, and I usually run it with 5min RI between sets, rather than 3min. I’ve still had to drop work intervals from the later sets, had to turn down the resistance, or just completely failed a set plenty of times!

When I’ve done this workout on my Wahoo Kickr I’ve prioritized a power target in Erg mode¬†and used HR as a descriptive measure of T90VO2max, aiming to maximize duration >88% HRmax as per¬†R√łnnestad & Hansen’s prescription (reminder, I now recommend >90% HRmax).

Prescribing Intervals via HR or Power

If my intervals are prescribed via target HR (eg. maintain >88%¬†>90% HRmax) I will naturally put out greater power at the beginning of the workout to rapidly bring my HR up to target. Once I’m at my target and as fatigue begins to accumulate, my power will end up decreasing through the workout and my HR will likely remain steady at or slightly above my target.

This naturally produces a decreasing-power interval, which I talk further about below.

If my intervals are prescribed via target Power (eg. work intervals at 400 W) I’m more likely to hold a constant power output through the workout, especially if I’m locked in to Erg mode. My HR will be slower to rise initially, but eventually may reach higher HRpeak than it would in the HR-prescriptive workout, since even as I fatigue the power will remain constant and unrelenting. Or I might even fail the workout at the given power target if I fatigue too rapidly.


  • Faster rise in HR, then plateau
  • Constant HR for remainder of the workout
  • Decreasing-power intervals


  • Slower rise in HR
  • Increasing HR through the workout
  • Higher HRpeak and potential failure
  • Constant power intervals

Decreasing-Power Intervals

I was very satisfied following the¬†R√łnnestad 30/15 protocol as my go-to VO2max workout. I felt like I was able to reach a good balance of output workload (power) and input workload (HR) in order to maximize T90VO2max.

Then I encountered some studies looking at decreasing-power intervals

Recall above that when prescribing workouts via HR, power will naturally decrease through the workout as fatigue increases and the heart either has to work harder to produce the same power, or work as hard and produce less power.

Decreasing-power intervals are where power is manually manipulated to maintain VO2 above target as long as possible Рie. to maximize T90VO2max. This is done with an initial severe intensity workload to rapidly raise VO2 and HR, then reducing power to match rising fatigue and maintain either VO2 (if measuring in a lab) or HR above target.

Billat et al, 2013¬†found that by carefully manipulating power workload in real-time responding to subjects’ measured VO2 (via gas exchange analysis), they could maintain >90%VO2max continuously in excess of 15 minutes!

One subject was able to last fully 26 minutes at >90% VO2max!!

Billat et al, 2013. Shows constant power interval (blue) where subject lasted ~8min >90% VO2max. Compared to decreasing-power interval (gold) where the same subject lasted ~26min >90% VO2max

This is obviously very compelling. But it would be difficult to manually manipulate target power in real-time for my athletes, or myself outside of the lab environment. So I would need a more generalizable prescriptive protocol to plug into a custom Erg mode workout to be able to follow at home.

Practical Prescription for Decreasing-Power Intervals

Another study, Lisb√īa et al, 2015¬†looked at a similar decreasing-power interval protocol. They used a standard method of starting the interval at the¬†maximum workload to elicit VO2max; what they called Ihigh, and what I call MAP, and finishing the decreasing-power interval at 110% FTP.

The interval duration was determined individually as the longest duration @ MAP (what they called Tlow, and is often called¬†Tmax… sorry for the acronym soup) This duration was then divided into 4 equal intervals of decreasing-power, from MAP/Ihigh to 110% FTP.

They found the sustainable duration >90% VO2max (T90VO2max) was over 50% higher, and total O2 consumption was over 70% greater in these decreasing-power intervals compared to constant power intervals at MAP @ Tmax.

Lisb√īa et al, 2015

Combining Intermittent Short Intervals and Decreasing-Power

My immediate interest after seeing these studies was to combine the two workout protocol to rapidly raise VO2 toward >90% VO2max (which can’t be measured at home) and HR toward >88%¬†>90% HRmax (which can be measured at home), and to sustain that duration for as long as possible.

I would start with a longer hard-start interval at MAP (highest workload over the shortest duration that elicit VO2max) then drop into a decreasing-power 30/15 protocol, finishing around 110% FTP.

T90VO2max… it hurts so good!

My HR rapidly reached >88% HRmax and then plateaued as the power declined at roughly the same rate at which fatigue accumulated. This particular workout I pretty much nailed the balance and managed to accumulate >36min @ >88% HRmax! (update: although the accumulated time was much less at >90% HRmax)

I was using 88% HRmax as my target based on the R√łnnestad & Hansen 2013 & 2015¬†studies, although I now believe a better target for most people is 90% HRmax. The big assumption here is that this HR target correlates to 90% VO2max… an assumption that I’ve come to challenge!

The good thing about using Erg mode in Zwift or TrainerRoad is that I can have my hand ‘on the trigger’ to turn up or down the resistance in real-time. If I saw my HR drifting away from my target, or if I felt like I was about to explode I would rapidly raise or drop the target power a few percentage, which would instantaneously adjust the resistance of my Wahoo Kickr.

WKO4 chart showing time and work (kJ) spent in approximate VO2max (orange) and Threshold (yellow) zones for the same T90VO2max workout as above. Note the decreasing power toward Threshold during later sets as fatigue accumulated. Despite lower power (output workload), my HR (input workload) remained above target >88% HRmax.

Prescribing T90VO2max Workout for Yourself

First, the disclaimers. Ultimately, there are a lot of assumptions built into this VO2max training protocol. Since we are unable to easily measure actual VO2 at home or out on the road we need to use correlated metrics to estimate our VO2 at any given time.

Namely, HR and Power.

There is an interdependent relationship between HR, Power, and VO2 that I still don’t fully understand.. but they have to be in the right balance for optimal adaptations to occur.

I believe targeting maximal duration >88% >90% HRmax should correlate well is the closest we can get to estimating duration >90% VO2max, and that T90VO2max is a critical parameter for enhancing VO2max and ultimately Aerobic fitness and endurance performance.

Ideally before prescribing this workout for yourself you should have a good sense of your MAP (Max Aerobic Power; the maximal power that will elicit VO2max) and the maximum duration at this workload (MAP @ Tmax), and of course your FTP. MAP can be estimated at home with a Ramp Test (which I will detail in a future post) or in a lab-based VO2max or lactate test.

Unfortunately, there’s no easy way to prescribe this workout based on percentage of FTP, since everyone will have a different relationship between MAP and FTP (see the recent Fast Talk Podcast for a fantastic discussion on the hyperbolic “death of FTP”).

Trainer Workout

The workout protocol I’ve been using is¬†3 sets of 13-15x 30/15s, 3-5min RI between sets. This should accumulate around ~10min of work each set, for ~30min cumulative work duration.

To do this workout on a smart/Erg trainer, the first hard-start interval should be 1 minute @ MAP, before going into the 30/15s work:rest short duration for 12-15 work intervals. Power will then decline in regular steps every ~3-4 work intervals toward ~110% FTP by the end of each set.

Your target will be to maintain HR above >88%¬†>90% HRmax. HRmax can be your highest observed HRpeak in the past 6 months, or I like to use the HRpeak observed during your most recent VO2max Ramp Test within the last 6 months. This avoids potential false data spikes and ‘race HR’ that may not be achievable during training.

The second and third sets can have shorter hard-start intervals, as HR will more easily rise toward your target due to accumulated fatigue, and spending too long at MAP will over-fatigue you for the remaining work intervals. I found average power for the later sets will inevitably be lower than the first set, but this is fine as long as HR remains on-target.

If possible, have your hand near the trigger to adjust the power up or down as necessary in your workout app of choice, to maintain HR on-target.

Outside Workout

If you were to do this workout on the road the¬†prescription would be simply ‘near-max effort’ for 1min hard-start, then monitor and maintain HR above >88%¬†>90% HRmax through the remaining 30/15 set. You would naturally put out more or less power through the workout in order to maintain HR just above 88%¬†90% HRmax. And again average power will likely decline for the later sets.

I haven’t actually tried this workout out on the road, so if anyone does.. tell me how it goes! And if you changed anything to make it easier/more realistic to complete.

Sample Workout Files

don’t make me regret posting these…¬†