Empirical Cycling Podcast “Why Not Rønnestad 30/15 Intervals” aka Brasted

Scientist do not always equal scientists. I am currently finishing my PhD and sometimes I am really terrified by some “science” papers. I know couple of respectable scientists that have completely no idea what they are talking about. But like you said - good scientist are open for a discussion and critique.

5 Likes

They tested Hunter Allen’s protocol, which includes a blow-out 5 minute effort as part of the warm-up.

Thanks for highlighting the new study. I can’t access the paper atm for some reason. I’d be interested how they determined PO at RCP? Did they account for MRT to translate ramp power to constant work rate?

Interesting findings. Definitely not what I would have expected.

“Maximal Ramp Test
All the tests were performed on the same cycle ergometer (SRM,
Jülich, Germany). The test started at 75 W (which was maintained
for 5 min), and thereafter, PO was increased following a ramp-like
protocol (ie, by 5 W for every 12 s up to average of 25 W·min−1),
while gas exchange data were collected continuously using a
breath-by-breath system (Ultima Series Medgraphics; Cardiorespiratory
Diagnostics, Saint Paul, MN). Participants were required to
keep pedaling cadence between 70 and 90 rpm, and the tests were
terminated upon volitional exhaustion and/or when subjects could
not maintain a cadence ≥70 rpm. The RCP was determined as
the work rate at which an increase in both ventilatory equivalents
for oxygen and carbon dioxide occurred together with a decrease
in end-tidal carbon dioxide”

IOW, no attempt to account for the expected lag.

ETA: What did you find unexpected?

1 Like

I got my hands on the paper and just quickly reviewed the methods. Haven’t read the entire paper so no hot take-downs yet :face_with_hand_over_mouth:

But yea, appears they did not account for mean response time (MRT) between muscle and lungs, which should be considered when comparing power from a ramp test to a constant work rate trial. Relevant intro to the phenomena below.

But the difference they found between PO @ RCP and FTP was still 50 + W, which I think is still greater than what could be attributed to the MRT and ramp-CWR translation.

Unexpected to me that (1) there was so little difference in 20min power between different warm-up conditions. And (2) that FTP was that much lower than PO @ RCP (although correcting MRT would bring those estimates closer). The studies I’ve encountered (certainly not comprehensive) comparing FTP to other threshold measurements usually find a closer relationship, albeit with large limits of agreement/high individual variability.

image

image

edit: A more comprehensive review on the topic of translating ramp VO2 to constant intensity:

1 Like

Thanks for the explanation. I hadn’t spotted that the 20min test immediatly followed the ramp test. That sounds horrible, hah!

And thanks for pointing my attention to the individual data points, instead of just looking at the average. This paints a very different picture of the results.
Just looking at the max power from the ramp test (Fig 1):

  • in the 30/15 group, 3 out of 7 participants increased their power, one decreased it, and 3 didn’t change. Of those that increased, 2 increased a lot, they added 40W! (within 3 weeks in trained athletes!)
  • in the 4x5 group, 3 out of 9 increased their power, 3 decreased, and 3 didn’t change.

That looks like the discussed mean changes in power are overly influenced by individual responses, for example the two big improvers in the 30/15 group. From the statistics section, it doesn’t look like they made any attempts to control outliers either. I don’t think they should have averaged over those results. The participant number is too low, and responses vary too much. The reviewers really should have pulled them about that. Discussing the mean changes between these groups means nothing.

3 Likes

Almost all cross-sectional training studies are underpowered, making them highly susceptible to either type 1 or type 2 errors as a result of high or low responders. Longitudinal studies are better, but are even harder to conduct, and carryover is always a concern. Your best bet is therefore to not base your training decisions on the results of just one or a couple of studies, but instead on general principles and meta-analyses, like these.

8 Likes

@old_but_not_dead_yet The “general principles” thing is what I tried to get across with the podcast series. It was described to me by a listener as “first principles”, which may be more the case. With longer studies, we run into other training issues like classifying distribution of training time in zone.

@splash this is another common issue with many studies like this. Statistics can be misleading, and it was nice of them to publish the individual responses, which is becoming more common these days but actually seemed to be more common in the 70s-90s. One other issue I forgot to mention is that if their thresholds are in the 330w range, this puts national level riders with an average bodyweight of 75kg at 4.4w/kg at FTP. That’s also way too low, so I expected 5-5.5w/kg which puts FTP minimum at 375w.

There’s another funny one I saw recently. Kolie Moore on Instagram: "How dare anyone fit trend lines to this data." The n is so large here, despite the basically nonexistent r or r^2, that when they compared the data sets against each other they ended up with p<0.01 and much, much lower.

To sympathize with the Ronnestad paper (and unfortunately the cut section of the podcast included this), it’s really hard to find participants for a study, let alone elite athletes who don’t mind someone screwing with their training program. His methodologies are actually pretty par for the course too, which is something that saddens me frequently when reading through the exercise phys lit. It’s also why I generally pick a good paper for the podcast and make my own conclusions from the results and compare those to what the authors thought. I would have ignored this paper completely were it not for its popularity, and the huge disconnect between the quality of the result, its lack of understanding of what’s “under the hood”, and the subsequent interpretation in training media and culture. And this is why I had some not nice things to say about Ronnestad and his understanding of physiology and training, which extends to trying to get a positive result by sweeping inconvenient data points under the statistical rug. And it’s par for the course too, and part of the reason I like older studies better is in part because they were addressing more fundamental things, but also report a lot more individual data with better methodologies. Biochem papers from the 40s have pages on pages of data tables.

Publication itself seems to be thought of, by the public at large and even some in the sciences, as a stamp of Official, Indisputable Truth. If I were a reviewer (not that there’s more than four people in academics who care what I think, and only one has any influence on anything) I would have done my best to improve the transparency of the paper since it would be basically impossible to get those athletes again in the same starting conditions. But I think I’ve said elsewhere that the real peer review is after publication, when the rest of us get hold of the paper and can find methodological improvements, maybe try to replicate the results, or just say they’re wrong… and I very much tried to be constructive by putting forward my understanding of what’s under the hood.

Now if you want to read a great exercise physiology paper, try this one. I really, really wanted to find a reason to include it in the vo2max series but couldn’t find a good enough reason to work it in. https://physoc.onlinelibrary.wiley.com/doi/full/10.1113/JP270408

1 Like

I am absorbing as much as I can of this discussion, though I really need to listen to the back episodes of the podcast to better understand the physiology, and am finding it very interesting - tangents and all. Now to put a slightly different slant on my original post let me put the following question:

Brasted - what is it actually good for, and for how long?

I may have adopted Brasted for the wrong reason but I like it’s 30/15 format, which is to say I don’t dread it in the way I would the 6x3min format of something like Spencer +2 for instance. It makes it repeatable in a way I would find hard to do with Spencer. Its not physically easy for me by any means - I do it at an intensity at which completion is touch and go and requires me to dig very deep. I do however find it more approachable mentally. So whilst it looks like I need to bite the bullet and find a more effective VO2max workout I would be happy to be given a good reason to keep Brasted in the mix.

1 Like

My personal reasoning, which is 100% anecdotal and 100% non-science, is for the reasons Kolie lists:

Doing 60-80 high RPM 30/15s in a workout really makes my hip-glute-leg chain feel more engaged and connected vs a quad busting 5x5 workout, especially if I’m coming out of an “off” period or even a long SS block. I find it doesn’t take many to build those sensations, maybe a couple of weeks. That’s just my perception, which very well may be wrong from what’s actually happening.

2 Likes

Yeah, agreed, at best they seemed to sort of skate around the issue, versus tackling it head on.

“20-minute TT used to determine FTP.
Our group recently observed a good agreement and a strong
correlation (r = .95) between FTP and the lactate determined with
the “D-max” method,4 with the former having been previously
reported to present a strong correlation (r = .97) and no differences
with the MLSS.7 Similarly, Borszcz et al5 reported no differences
and a significant correlation (r = .61) between FTP and the individual
anaerobic threshold. These authors observed that the
FTP-associated PO could be maintained for ~51 minutes before
exhaustion,5 and indeed, the time to exhaustion at the MLSS has
also been reported to average ∼50 to 60 minutes.8 Confirming
these results, the same authors recently reported no differences and
a strong correlation (r = .91) between FTP and MLSS. In the
present study, the FTP was lower than the PO eliciting the RCP,
and although there is controversy,1 some authors have reported
that the MLSS also occurs at a lower PO than the RCP.2 However,
we did not analyze the relationship between FTP and MLSS, and
we did not assess if exercising continuously at the FTP or RCP
resulted in a steady physiological state

1 Like

going to listen to the podcast today.

n=1, but I’ve seen better results from 4-6min continuous intervals than from Tabata or other forms of intermittent work. My W’ PMax, whatever, is low, and I just can’t get much out for :20-:40. When I do Tabatas, my HR looks like a continuous interval but the power is 20-30w lower. Just bang a 4-6min hill or flat stretch, for me.

1 Like

For me, that’s one of the big draws of VO2max – individuality. We all have very similar but also very different physiologies; trying to suss out what those differences are and which training will lead to the greatest/best enhancements is half the fun of ridin’ bikes. Especially for the self-coached, we have to be both Sherlock and Watson at the same time!

2 Likes

Here’s a webinar by Dean Golich on doing max aerobic intervals. He advocates going max on them, even if it means power might decline throughout the set.

3 Likes

Hmmm. Interesting, but I am old and set in my ways. Now, my VO2 work on the trainer is all high cadence, since I have a fixed gear with a powermeter on a Kinetic Road Machine.

On the flats, sure, 100+ rpm is the only way I can make the power (and isn’t the goal to make the power and move the dang bike?). On my interval hill, well…the best way for me to make the power is to stand for the first 90 seconds, grinding the big ring at 60rpm. There’s a 90rpm false flat, then another stand and grind, then I’ll hit the last minute of a 5min interval at 100+ on the flat after the hill. Not what Kolie says is ideal, but I’m still going to do it that way.

I’m one of those low AC/FRC guys – so I just get up that hill as fast as I can, then hang on for another minute or two after the crest. I do start fast – about 425w for the first minute (the 5min intervals I did today were at 385avg…FTP is maybe 335-340 right now…I don’t test…I just get up the hill as fast as I can).

I don’t see high heart rates after the first couple of interval workouts in a cycle. The watts for those 5min efforts will go up by 10-15 over 6 weeks, but HR drops from an average of 154 (my 90% is 151) to an average of 148-150- so maybe that stroke volume improvement is happening anyway.

Now that I have my own erg, I may do some Transport intervals on it – there’s no way I can pull splits at my 2k pace without 32+ strokes per minute, so maybe that pre-loading will happen with my heart on those buggers, even if I’m still doing to stand and grind on the hill repeats.

1 Like

Finally found time this morning to listen to Mikael’s interview with Michael Rosenblat in the most recent TTS podcast (I do love Mikael’s podcasts just hadn’t got to it yet). It was very interesting to hear what he had to say on what might be happening in what he referred to as ‘short’ HIT interval workouts - the domain of the Ronnestad/Brasted format. As you point out he is distinguishing between positive adaptations in VO2max on one hand and MAP on the other. It gets to the point of considering not just how much oxygen uptake increases but also what that extra oxygen might actually be doing for you - the locomotion vs phosphocreatine replenishment point. The conclusion being that if you are looking for VO2max improvements associated with improvements in MAP and time trial type performance, then the longer VO2max interval format is the way to go.

For anyone with just a couple of minutes to dip into this TTS podcast to get a better sense of what I’m trying to summarise here, just drop in at the 23min point for a couple of minutes or so.

3 Likes

I would normally not say anything here, but since we’re on a thread that started with my podcast… Raising MAP but not vo2max is most likely improvement in anaerobic capacity as we discussed in WD#18. Rosenblat (so far) makes no mention of anaerobic capacity or metabolic buffering capacity but he does mention recruiting larger motor units (specifically as “fast twitch fibers”). Kyle and I did mention PCr replenishment by mitochondria in WD#2 but did not actually talk about greater motor unit recruitment, which I hesitated to do so early in the podcast but will be getting a lot of attention at some point.

I do greatly appreciate his distinction of the 2 minute time range for intervals. I find vo2 kinetics are actually a bit faster than that but it depends on how hard you start the work bout. In the TTS podcast (probably my favorite of all the training podcasts out there, if I had to pick one) Rosenblat also doesn’t get deep into variation of rest intervals (or my brain wandered a bit), since a 30 second rest interval between 30 second all out sprint efforts is a much different stimulus than a 4-5 minute rest interval, which is how some SIT studies are commonly conducted. So I checked through the actual paper, and indeed the SIT rest periods are ~4 minutes and greater. HIT intervals are varying between intermittent and continuous somewhere between 80% and 120% of “MAP”, which is in itself an eyebrow raiser because it doesn’t take into account where it is relative to LT, which is pretty typical of the exercise physiology literature but, again, I think needs to be addressed.

Anyway, replenishing PCr is locomotion. PCr is a spatial and temporal phosphate buffer, meaning that its greater diffusion speed and capacity is used to move energy out of mitochondria faster and more easily than actually moving ATP. There are various papers looking at conveyor-belting of ATP and PCr, and the structure and stability of the mtCK complex. So no matter what, more mitochondria is more PCr repletion at any exercise intensity. But we also need to take consideration for motor unit recruitment, anaerobic capacity, buffering capacity, glycogen stores, etc etc.

So when it comes to comparing what’s looked at in this study, there’s no actual difference between HIT interval lengths or intensities. All it’s really showing is that working continuously or intermittently with short rests is going to have a greater impact than the odd all out sprint with “long” rests. Which is like… duh.

2 Likes

Can someone sum up / suggest the interval training best suited to raise vo2 max?

3 Likes

Overlaying confidence interval lines on regressions is usually pretty insightful - CI for the mean, and individual predictions. If they did that for the data you linked to, the CI bands would be so wide where is would be clear the data isn’t really that useful. And in this case, individual variation far overpowers the average trend.

I have a long drive ahead today, and have your VO2max podcasts on the play list.

So with that as an acknowledgement - that I haven’t yet listened to your VO2max podcasts - I’m hoping I’ll gain some insight into answering the questions below.

If there’s anything you’d add to your podcast content re. the questions below, I’d be interested to hear.

Cheers,

Dave.

  1. What are the factors that drive MAP (note: I guess I could say VO2max instead of MAP, but I’ll say MAP as that’s what moves the pedals).
    A. Max cardiac output (SV*BPM)
    B. Oxygen carrying capacity of the blood (quantity and quality of red blood cells; blood volume - and presumably this also influences max cardiac output)
    C. Ability of working muscles to extract oxygen from the blood
    D. Efficiency with which oxygen is utilized by working muscles to create mechanical power
    Note: from what I understand, right ventricle cardiac output, and ability of lungs to oxygenate blood is not typically a limiting factor - this accurate?

  2. How can I tell which one(s) are my limiters?

  3. What kind of training can move the needle on my limiters?

  4. Is knowing the answer to #2 and #3 even helpful - or is it the same kind of training no matter what?