Polarized or Pyramidal Training for Ultrarunning with Michael Rosenblat, PhD #238

Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
Speaker 1 (00:08):
Trail and ultra runners.
What is going on?
Welcome to finally.
Welcome to another episode ofthe coop cast.
As always, I am your host,coach jason coop, and it has
been a while for new episodes ofthis podcast, but we are back
in action and I am happy toannounce in a slightly new
format More on that later,because today's guest is making

(00:31):
his second appearance on thepodcast.
Welcome back, michael Rosenblatt, who recently was the lead
author on a new paper ontraining intensity distribution,
which gets to the age oldquestion of what types of
interval workouts over longperiods of time are the most
effective.
Michael comes back on thepodcast to discuss his work,

(00:52):
some of the painstaking steps hehad to take to collect the data
, as well as what all this meansfor the upcoming workouts that
you have in your calendars overthe course of the next weeks or
months.
And finally, as a new wrinklein this podcast, we'll be
hearing from our CTS coaches atvarious points during the
discussion to bring a practicalelement to the research and how

(01:14):
we use it with our athletes.
So today on the podcast, youwill also be hearing takes from
CTS coaches Ryan Anderson andAdam Ferdinandson as we discuss
how we use these findings inpractice with our real athletes.
All right, we're back in thesaddle and, as always, I'm going
to get right out of the way.
Here is my second conversationwith Michael Rosenblatt, all

(01:37):
about intensity distribution andprogram design.
Hey, how you doing?
You're doing pretty well.

Speaker 2 (01:45):
I'm glad we could finally get this going yeah,
there's a few little hiccupsalong the way I can see like I'm
just in a random airbnb now.

Speaker 1 (01:53):
So anyway, what was me anyway?
Thanks for putting up on myschedule.

Speaker 2 (01:56):
Yeah, no, it's great that we're able to to meet up
and discuss the new research.
It's kind of exciting stuff.

Speaker 1 (02:03):
So yeah, yeah, I'm stoked about it too, anytime we
can talk about training,intensity distribution and
things that coaches willactually take away and apply
right, because that'sfundamentally what we do is
prescribed training, those.

Speaker 2 (02:17):
That's usually the research that we're kind of the
most keenly interested in yeah,and I don't remember if I
actually explained that to youlast time that we spoke, but
that was actually why I got intoresearch, because I was
originally a coach, yeah, right,so I coached triathlon and so
it was to find better ways totell my athletes and I felt this
was the best way to go.
So, yeah, that was really tofind, you know, do good work and

(02:41):
more applied work.
So there's a lot of theoreticalexercise science stuff out
there, but it's good to be ableto or hopefully it's good to be
able to find some good work andcreate some new knowledge
that'll be directly applied tothe field.

Speaker 1 (02:55):
Yeah, that's good.
I mean, we might as well getinto it, we don't?
We can skip, skip right to it.
Well, welcome back to thepodcast.
First off, I had to go and lookback.
The last time you were on thepodcast I released that episode
and let's see, when was this?
I believe it was December of2013.
So a little over a year ago, orsorry, 2023.

(03:16):
I'm even going back evenfurther than that, Number 198.
So it was a while ago, and howtime flies and I'll leave a link
in the show notes to thatparticular podcast, but it is
actually a pretty decent primerfor this one, because we're
talking about some of thepractical elements that we're
going to get into here, which isjust how should we design

(03:36):
intervals right, like, from afundamental standpoint.
You want to do something with anathlete, you want to improve
their VO2 max or improve their40K time trial or whatever it is
?
How do you design theindividual training sessions to
accentuate either a particularpart of their performance or a
particular part of theirphysiology, or some Venn diagram
overlap between those two?

(03:58):
As we were just discussing,those types of conversations end
up being some of my favoriteand also some of the listeners'
favorites as well, because theyend up going and doing the
things that they just heardabout sometimes the same day,
sometimes the next day,sometimes the next week, and so
I'm always really, reallythankful of that.
So you kind of already gave alittle bit of a background on
like how you started to answerthese questions, but if you want

(04:20):
to elaborate on that a littlebit more, just in terms of like
who you are, your coaching andyour research and things like
that, go ahead and take thefloor for a little bit so the
listeners can get to know you alittle bit better again.

Speaker 2 (04:30):
Sure.
So originally I was actually atriathlon coach.
I think I started coaching inmy early 20s and I'm from
Toronto, ontario, in Canada, andI think there's maybe four or
five other provincial coachesaround and I started working
with them and learning along theway, and I was asking a lot of
questions about you know bestways to program exercise, be it

(04:51):
interval training or even goingup to the full program and I
feel the answers I used to getwas well, this is what's been
done in the past and so this iswhy we do it this way, or this
is what someone else has showedme, which I mean.
Of course, clinical orpractical experience is
certainly valuable, but I kindof needed something a little bit

(05:12):
deeper than that, and so that'swhy I continued my education,
and even actually before Icontinued my education, I
started reading the literature.
When I say continued myeducation, I mean formally, but
I'm actually a physiotherapistby trade.
So after coaching triathlon, Igot into physiotherapy and after
practicing for a little while,I decided to go back to do a PhD
in exercise science at theUniversity of Toronto, and my
research was primarily inprogramming interval training.

(05:35):
And, of course, covid happened,and that shifted my research
focus from laboratory work tomore knowledge, evidence
synthesis and knowledgetranslation.
So I got to start synthesizingthe evidence and I think I
became somewhat of amethodologist and biostatistic
along the way, and that's kindof what the majority of my work
has been, or the area of my workhas been.

(05:56):
I'm currently practicingfull-time as a physiotherapist,
but I'm also a researchaffiliate at the Sylvan Adams
Sport Institute and kind ofguiding, hoping to help further
their research program as well,and yeah, and so that's kind of
where we are today and just tohave this new research that's
coming out.
It's quite an interestingproject because of the

(06:17):
collaborative nature of the work.
Yeah.

Speaker 1 (06:20):
So that's what we're going to talk about today.
It's a synthesis of theinformation, as you put it.
I like that terminology andyou're absolutely right.
It is a kind of a hit list ofwho's who in physiology.
A lot of these people have beenon this podcast, including
yourself.
I've got to get Jim Arnold onbecause I just really respect
his content that he puts out inthe space, but I think five of

(06:43):
the authors in this list havebeen on this podcast either one
or multiple times, and what I'mgoing to do in the show notes,
since a lot of these themes areredundant, is link up those
podcasts as well.
So if somebody wants to do areally deep dive into all this
stuff, which has a lot ofoverlapping topics with them, I
would encourage you guys to justlist in sequence but the

(07:05):
specific piece of research thatwe're going to talk about now,
the title of which is and if youhave changed the title since
the manuscript that you sentover to me, please correct me,
but it's a mouthful the linkwill be in the show notes.
But the title is which trainingintensity, distribution,
intervention will produce thegreatest improvements in maximal
oxygen uptake and time trialperformance and endurance
athletes?

(07:26):
And to say it's a synthesis ofthe information I think kind of
like under undersells the effortthat you and your colleagues
undertook and we're going tokind of get into that effort
just to paint the picture.
The title kind of says it allbut a little bit.
Elaborate a little bit on likethe questions that you're
specifically trying to answerhere.

Speaker 2 (07:46):
Yeah.
So it's actually funny because,even just to discuss the title,
that was certainly a discussionamongst my colleagues,
especially because it's quite along title and we tend not to
like to have those types oftitles but, based on the type of
work it is, there's certaincriteria that we need to follow
in the title.
So it's kind of as much as it'slong, it's also somewhat
necessary, unfortunately.

(08:06):
But yeah, in terms of thequestion that we wanted to
address was looking out, youknow, I'd say well, which
training intensity distributionwould be the best for an athlete
to use to maximize theirperformance.
And interestingly, you know, weadded the term intervention
after the training intensitydistribution, because when you

(08:27):
really think about research andsports science, we're really
just looking at an interventionin a set given amount of time,
whereas in the real world we'rereally looking at not just, you
know, a single session or a week, or we're looking at a full
program to find a way to getsome an athlete to peak or to be
able to peak multiple timesthroughout a season.
And so, yeah, and so I thoughtwe kind of realized it was

(08:48):
probably the best way todescribe what the work is that
we're doing.
It's really just looking at anintervention period and the
questions that we wanted to lookat.
You know we use those twovariables or outcomes maximal
oxygen consumption and timetrial.
Maximal oxygen consumption iscertainly probably the most
common physiological variable orinternal measure that we use to

(09:09):
measure performance, and thentime trial, I'd argue, might be
the best.
It's hard to say, I mean Imight say it's the best
measurement of performance interms of specificity.
It's as close as we can get towhat an athlete might actually
be doing.
But of course both have theirstrengths and both have their
limitations, and there's so muchto this that we're trying to

(09:32):
address.
But I think there's a fewreally unique things about that,
and one, unlike previousliterature that's been done,
including my own work and Ithink it was some of the
foundation of what we werediscussing in our first time in
our first podcast was it wasn'tjust a pairwise analysis,
meaning it wasn't just comparingone intervention to another,
it's a network meta-analysis.
So we were able to compare allthe different types of training

(09:55):
intensity distributions to eachother, and so that's something
that's unique about the workthat we were doing no-transcript

(10:31):
.

Speaker 1 (10:31):
a coach, I mean and you know, you know this from
your own coaching practice oneof the things that is the
biggest time cost when you'reworking with any athlete, and in
particular any new athlete, isdigesting their previous
training, because it'severywhere it's in Strava and in
training peaks and on the backof napkins and in three ring

(10:52):
binders and you know old Excelsheets, and sometimes all of the
above with just one athlete andbeing able to come through that
for any significant period oftime is always a pretty it's
always a pretty big time cost.
So when you and I were offlinekind of discussing how we're
going to crack this conversation, it kind of like to for you to
peel the curtain back a littlebit because I think it just it's

(11:14):
kind of speaks to the effortsyou have to go to to like figure
this stuff out.

Speaker 2 (11:18):
Yeah, and not only that, but to find a consistent
variable in terms of theirtraining data that goes across
all of the studies.
Of course you can use differentvariables and try to mesh that
all together, but you'd want tofind something that's a similar
measurement, to be able toaccurately and I guess I can say
reliably combine the data.

Speaker 1 (11:36):
So you're looking at all these different studies,
right?
Can you kind of give a littlebit of the range of the studies
that you were looking at interms of how long the
interventions were, what thetypical interventions look like,
who the participants were andthings like that?
Just so, when you're digestingall of these different studies,
what I'm trying to communicateto the listening audience is is

(11:58):
what's kind of the size andscope and what do those studies
actually look like that you'relooking at in terms of how
relevant it might actually be tothem?

Speaker 2 (12:07):
Sure.
So first of all, we onlyincluded experimental studies or
quasi-experimental studies,meaning like they're randomized
into different groups or atleast somehow put into two
separate groups, and that's howwe compare those different
groups or interventions.
The studies, the durations ofthe studies were quite different
, so some were as short as threeweeks, others were as long as

(12:27):
23 weeks or a couple of months,almost like six months worth of
training.
Sample sizes, you know,anywhere from maybe 10 people in
a study, depending on if it's acrossover design up to about 30
or 40 participants.
So sample size is not thatlarge, and I'd say that really

(12:48):
makes things interesting whenyou're looking at combining the
data.
Look at all sorts of sports.
I think the majority of sportswas running.
I think most studies includedrunners.
There was some with cycling,triathlon, cross-country skiing,
and in terms of performancelevel, they were all endurance
trained athletes.
But there was a kind of adichotomy here of athletes that

(13:09):
were either consideredrecreational or competitive, and
so we were able to do severalanalysis based on the different
characteristics of theparticipants included in our
studies.

Speaker 1 (13:21):
Yeah, and how would you more colloquially define
that dichotomy of a recreationalathlete versus a competitive
athlete?
Like, like, make it real to theperson listening and who's
going to be in one categoryversus who's going to be in the
other category?
Sure.

Speaker 2 (13:36):
So it's interesting because what we originally did
was we used the subjectivedescription that each of the
respective authors used, but wefound something that was
consistent across those studies.
So recreational athlete we'vecalled somebody who is just
doing a sport, specifically, sorunners who are running
consistently for a certainperiod of time.
You know, we can say, let's say, two months, three months, six

(13:59):
months, and they're training two, three days a week, but they're
not necessarily competing,whereas we're looking at
competitive athletes.
We say somebody who might be atier one athlete, a national or
provincial state level athlete,olympic level athlete.
So while there's certainly arange of competitiveness across
even that group of individuals,we found just those who are

(14:21):
technically competing at somedegree level that's been
classified previously versusthose that haven't been.
And that's how we originallyseparated the groups.
It's very common to separateindividuals by their VO2 max and
say, okay, well, we have thesepeople that are in a certain VO2
max and others that don't, andwe're going to slip them right

(14:42):
down the middle or come up withsome sort of arbitrary value.
And there's certainly an issuewith that, because when it's a
continuous outcome measure, howdo you know where the right
number would be?
And then the other issue withsomething like that would be
well, and I'm sure most of yourlisteners know that VO2max isn't
really the be-all, end-allright.
Submaximal performance isreally something that can

(15:03):
influence performance.
So, yeah, we use basicallythose subjective ways to
differentiate.
However, what was interestingwas we did a post hoc analysis.
Basically after we did all theanalyses, I looked at the VO2
max for those different groupsand they were statistically

(15:23):
different, which is quiteinteresting.
So it actually says, hey, maybethere's something to the way in
which we categorize theseindividuals.
So the recreational athletesthey had a VO2 max of roughly 55
mils, I think, plus or minusfive mils.
And the competitive athletesthey were around 65 mils, plus
or minus five mils, and includedboth males and females in those

(15:46):
categories.
So there's why we might have aneven larger standard deviation.

Speaker 1 (15:50):
So you ended up categorizing the recreational
versus the competitive athletes,kind of along their VO2 max
versus the whatever they werecategorized previously.
Am I understanding thatcorrectly?

Speaker 2 (16:02):
Well, no, actually what it is.
We didn't necessarily categorizethem by their vo2 max, they
just so happened to just backedit up essentially yeah, we
looked it up retrospectively wesaid, hey, it's interesting that
this subjective way in which wedecided to to separate these
athletes also showed somecharacteristics that were very
different and their vo2 maxeswere very different.

(16:23):
So it was kind of aninteresting finding which sounds
somewhat intuitive.
But it's nice when you do itbackwards and you see, oh, it
just so happens that it's moreof a real differentiation rather
than some arbitrary reason toseparate their VO2 maxes.

Speaker 1 (16:36):
I'm going to pick up on that backwards theme a little
bit because I was just thinkingabout this.
So we've already described theoutcomes that you were looking
at time trial, performance andVO2 max.
Then we've looked at thesubjects.
They're either recreational orcompetitive.
The interventions is kind ofthe first piece of it right.
So polarized versus and we'llget into a coaching discussion.

(16:57):
I'm sure that it's really not aversus thing once you get into
actually coaching athletes,especially when you look at it
over long periods of time.
But polarized versus pyramidaltraining and we've discussed
this across a few differentguests on this podcast.
But for the people that are new, let's just describe that
really quick, just so they knowwhat types of intervals and what
types of work kind of goes ineach one of those categories.

Speaker 2 (17:20):
Sure.
So first I think it's importantto describe that there's these
different zones of training andof course there's different
models that people could use.
In physiology we typically usea four or domain model In the
training intensity distributionresearch we typically divide to
three zones below your firstlactic or ventilatory threshold.

(17:40):
So basically very easy workbefore you start really to see
some sort of accumulation oflactate in your blood or before
your ventilation really startsincreasing.
And then you have your zone two, which is I guess we'd say is
really your threshold zone, andso maybe your race pace or
racing should be somewhere inthat zone two, depending on the
distance of course, but that'swhere you'll start to see an

(18:02):
increase in lactate and oxygenconsumption, but you'll still
reach a steady state.
And then zone three that'swhere you're above your maximal
metabolic steady state.
So now you can't reach a steadystate, you can only stay there
for a certain given amount oftime, of course, theoretically,
and at some point in that zoneor domain you will reach your
VO2 max if you stay there longenough.

(18:22):
And so if you think about it,that first domain, that's your
easy effort kind of training.
The middle zone is yourthreshold somewhere.
Racing sometimes typically isthere.
And then above that's whereinterval training would be,
anywhere from long durationintervals all the way up to
sprint intervals.
And so when we look at thedifferent training intensity
distributions, I think the bigone that we talk about a lot is

(18:44):
polarized training, and that'swhere majority of the time you'd
say maybe roughly 80% of thetime would be in that first zone
or domain and then 15, 20% ofthat time would be in the third
zone there, and then maybe alittle bit of time in that
middle zone.
And then we have our thresholdintensity distribution which is
I think that's the next, that'sreally where the two biggest

(19:07):
comparisons have beenhistorically where majority of
the training is in that kind ofthat middle zone there, with the
next next amount either in thezone one or zone two sorry, zone
one or zone three.
And then there's a thirdtraining intensity distribution
called a pure middle model,which is actually a very common
model.
I think it's something that'sbeen done quite a bit in the

(19:28):
past, but we only kind of moresay more recently, described it
as a pure middle model where themajority of training would be
in zone one, then the next wouldbe in zone two and then
subsequently zone three.
And so the reason why I saykind of we've only recently
started discussing it or callingit pure middles because a lot
of times we used to callthreshold training Right yeah,
we used to call it a thresholdmodel, when it actually really

(19:50):
maybe would come up with more ofa pure middle distribution.

Speaker 1 (19:52):
Yeah, and I've been.
It's funny that you mentionedthe threshold model.
So I've been coaching longenough to kind of remember some
of the original premise behindthat type of training, which
kind of alluded to the fact thatwe would call that the most
trainable aspect of an athlete'sphysiology.
We used to say the mosttrainable system, but we've kind
of moved away from thatvocabulary as well.

(20:14):
And because it was the mosttrainable part of an athlete's
physiology, well, naturally youwould have to train at that
intensity in order to facilitateall the adaptations and that
all the adaptations that arefundamental to performing at
that, at that intensity, inorder to facilitate all the
adaptations and that all theadaptations that are fundamental
to performing at that, at thatintensity.
Now, that's a, that's a historylesson, not what we're actually
going to, not what we'reactually going to go to.

(20:34):
But I do the newer vocabularybecause I think it starts to
move.
It starts to move away fromthat legacy and this like
pigeonhole ideology that youhave to train at these certain
intensities to elicit theadaptations of those intensities
.

Speaker 2 (20:49):
Yeah, it's interesting because we start to
think about what's thephysiological response that
you're trying to achieve versuswhat's the performance outcome
that you're also trying toachieve, and they're very
different in terms of how you'retrying to train.

Speaker 1 (21:00):
Yeah, Okay, so we've got these.
We've got these differenttraining intensity distribution
models.
How did you figure that outacross all of the studies?
Because there's a number ofdifferent ways that you can do
it.
If it's a cyclist, you can usepower meter.
If it's a runner, you can usepace.
All endurance athletes can useheart rate.
What were you fundamentallylike boiling down?
How were you fundamentallyboiling down the intensity that

(21:22):
the athletes were doing duringthese training interventions?

Speaker 2 (21:25):
So that's a very good question, and we actually we
did this in two ways.
And so there's and we'vediscussed this previously as
well, or you and I have spokenabout this a little bit in terms
of something called anintention to treat analysis
versus a per protocol analysis,and so the first way that we
analyzed how they did theirtraining was we kept the

(21:46):
athletes in their originalgroups.
So we said, hey well, the waythat the study was designed was,
they said, this group is doinga polarized intervention, this
group is doing a thresholdintervention.
Let's just say this is what itis and this is what their
intended distribution wassupposed to be.
So, based on what the originalresearcher for those respective

(22:06):
studies had stated, that was thefirst analysis.
And the second analysis, whatwe did was we were able to
collect all of the heart ratedata.
For when I say we, all of thecollaborators on the study were
able to look at all of the heartrate data that was collected
through their entireintervention.

(22:26):
And what they did was theylooked at the time in zone,
based on heart rate.
So there's several ways thatyou can.
Of course, you can measureworkload, as you mentioned some
of them.
You can look at power on a bike, you can look at speed with
running, and then there'sdifferent ways in which you can
calculate that doing, you know,looking at both internal and

(22:47):
external measures.
In this study specifically,heart rate was the only
consistent measure that we had,and not just that we had, that
would be consistent acrossstudies, but just that was
consistently, I guess, collected.
Because some of the studieswere a little bit older as well.
I think we have one study from07.
And while that doesn't soundthat old way, the way to conduct

(23:07):
research at the time maybe wewere just using heart rate
monitors.
Of course it was in runners aswell, and it was just at the
time, the best way to collectthe data.
I mean not necessarily for thatstudy, but I'm just saying
overall, heart rate was justvery commonly used, and so that

(23:28):
was what was consistent acrossstudies.
So we look at time and zoneusing at a given heart rate and
then, given that time and zone,we're able to determine how much
time each individual athletespent overall throughout the
intervention.
And in several studies we'reactually able to look week by
week as well.

Speaker 1 (23:40):
So here's my bonus question with that If you're
looking at each individual studyand trying to and using heart
rate for all of them to create astandardized intensity approach
, you also have to have a way todetermine what each individual
participants ranges were basedoff of heart rate.

(24:00):
And, as you and I both know,establishing that on the front
end with any one individualstudy can go several different
ways.
Sometimes they do, you know aramp protocol, vo2 max test.
Sometimes they're takingtraining data.
Sometimes they're doing youknow this, that and the other.
How did you stand?
How did you not standardize?
But how did you figure out thatcomponent?

(24:21):
Because you have all of theseindividual.
You have all of theseindividual athletes that are
doing different interventions.
You have heart rate data on allof them, but in order to
categorize the intensity, youhave to first calibrate what
heart rate means what zone foreach athlete.
So how did you go about doingthat?

Speaker 2 (24:40):
Yeah, and actually it gets even messier than that.
To make this more interesting.
The way in which the individualauthors determined the
different zones was alsodifferent.
And I don't just mean did onegroup use lactate, did one group
use ventilatory thresholds?
But some groups wouldn't useinflection points.
They would say, well, we'regoing to look at this.

(25:02):
You know, two millimoles oflactate would be their first and
four would be their second, ortwo plus one millimole would be
consumed, and so they were verydifferent across the studies.
And what you do, if you're ableto synthesize the data properly
and you're looking at, you'rerunning a true meta-analysis
here you do something called theOxford approach, which is where

(25:23):
you combine all the studiesfirst, regardless of the fact
that, hey, we know there's thesedifferences, maybe in the
participant characteristics, andin this case we're talking
about how we would determinethose different thresholds and
then determine, after we'vepooled that data, did that
actually influence the results?
And so it sounds like, well, ofcourse it's going to influence

(25:43):
the results, right, and so itsounds like, well, of course
it's going to influence theresults, right.
I mean, these are all thesedifferent measurements that are
different methods to determinethe same thresholds in terms of,
you know, are we using gas, arewe using blood measurements.
Where are we taking that bloodmeasurement?
Is it from the ear, is it fromthe finger, is it arterial, is
it venous, those types of things.
And so you know there's allthese different ways to do that.

(26:06):
And so you'd say, well,theoretically that sounds like
for sure, this is just a bigmess.
How can we pool all this datatogether and actually get
something that's consistentacross all studies to be able to
pull the data?
Now you know, if you're lookingat an individual participant and
you're going to say, well, Iwant to know, am I seeing a
change over time?
Then you're going to say, well,I want to know, am I seeing a
change over time?
Then you would need to say, yes, I need to be very consistent

(26:28):
with how I'm measuring this data.
And the reason why I'm sayingthat is because I don't want
this to be misinterpreted assaying, oh well, it doesn't
really matter which way we dothis.
A hundred percent it does.
But when we actually pooled allthe data together, there was no
heterogeneity or what we'll saya statistical heterogeneity
across all the studies when welook at the results, meaning
that the variability was sominimal that it actually didn't

(26:51):
matter when we pooled the data.
Okay.
So it's very important to say,though, while it may not have
mattered when we're looking at alittle bit of difference, when
you're training, it can be maybea little bit below, a little
bit above it's not going toinfluence you that much Maybe
your training it can be maybe alittle bit below, a little bit
above, it's not going toinfluence you that much maybe.
But when you're looking forchange over time in an

(27:12):
individual athlete, then ofcourse it definitely matters.
Right, if you're going toassess it from the beginning of
an intervention to the end of anintervention, then yes, you
need to be very consistent withthose measurements.
But overall, a little bit ofvariability here or there
actually didn't influence theoutcomes.

Speaker 1 (27:26):
You know, what's interesting is like the inverse
of that is normally what I haveto deal with as a coach, when I
have an athlete that does somesort of graded exercise test in
one lab and then, for whateverreason, wants to do it in
another lab with anotherprotocol, marrying those two up
in terms of were they better,worse or the same during those,
both those points when theprotocol is different, when the

(27:50):
equipment was different, thereare a whole number.
Even the technicians doing the,doing the tests are different.
In many cases I won't say allcases, but in many cases it's
hard to reconcile because youdon't know if this, if a test
that uses three minute stagesversus a test that uses four
minute stages and you see theinflection point at, you know X

(28:10):
minutes per mile or kilometersper hour or whatever, you don't
know if that's, if that has thesame meaning when the, when the
protocols are fundamentallydifferent.
But what you're saying is thatwhen you pool all of that, it
kind of doesn't matter, likethose differences, kind of like
they're not as I don't know whatthe right word I'm searching
for, but they tend to.
It just tends to not matter, Iguess.

Speaker 2 (28:32):
Yeah, but you're still onto something important
here, though, right, like ifyou're going to be training and
let's say you're training, youknow five Watts below or five
Watts above your threshold, well, I mean, physiology is kind of
continuous.
Everything's kind of alwayshappening at once, and to what
degree, and if you go a littlebit less or a little bit more,
well it's not at least theresults of this study suggest
maybe it's not going to havethat big of an influence.

(28:53):
But what you were saying,though, is if you're going to
send an athlete to one lab to doa test, even if it's the same
test but it's a different personinterpreting yeah, that
definitely would be a problem,because now you can't tell if
someone's improving, and that'swhere there's a problem for sure
.
So you're correct to say that,yeah.

Speaker 1 (29:12):
My colleague, Lindsay Golich, over at the training
center, who I've known for years, is really famous for saying
and I don't know if she's the,if she can claim origin over
this quote, but I'm going togive her origin over it that we
will always take a consistenttest over the perfect test
Meaning we get.
Sometimes we get caught up inokay, we, in order to do this

(29:32):
threshold test, we want exactlythese stages and exactly these
speed ramps and on, and then 10years later we come up with some
other version of that becauseof some nuance that somebody
thinks is important.
And what we'll always come backto is the test might not be
perfect, but if you do itconsistently, you're going to be
able to extract moreinformation out of what the

(29:54):
change over time is, becauseyou're not having to also
interpret what the change in theprotocol is and how that's
affecting the athlete and theprotocol is and how that's
affecting the athlete.

Speaker 2 (30:03):
Yeah, I like to think of it as if you're getting on a
scale and it's five pounds offright and you want to see if
your weight is changing.
Well, it's always going to bethe same five pounds off right
and so you'll know are you goingup or you're going down, but
it's always going to beconsistent there.

Speaker 1 (30:17):
Yeah, perfect, okay.
So let's get down to the nittygritty here.
So we've got these two types oftraining right Threshold
training and VO2 max training,or polarized training.
And sorry, pyramidal trainingand polarized training.
I want to make sure I've gotparallel structure there, sure,
two categories of athletes therecreational and the competitive
athletes and then the pool setof all of them and then two

(30:39):
different outcomes time trialperformance and improvements in
VO2 max.
It's two by two.
If you want to think about itlike that, let's use the
broadest lens possible, right,sure?
Did any of the intervention,either of the intervention
categories, affect either endperformance and time trial
performance or VO2 max?

Speaker 2 (30:57):
when we're looking at the groups as a whole, so we're
looking at it as a whole andthis also includes other groups
as well.
We look at differentdistributions that just so
happen to be included.
So I think there's a total offive different types of
interventions.
One of, the one of the groupsshowed no, there was no
difference at all.
And the other one showed nostatistically significant
difference.

(31:17):
And I'm very particular abouthow I say that.
Okay, that's very important tosay it that way.
And so when we look atpolarized versus pyramidal,
there was no difference betweenthe groups.
When we again we're just lookingat everything as a whole right,
I said I'd use this idea orthis concept of I can't at the
moment it's eluding me.
I've already said it the Oxfordmodel, basically how a look at

(31:40):
the results and then kind offigure out everything after that
.
So there's no differencebetween polarized and pure
middle.
We look at everything together,but there's no statistically
significant difference acrossany of the other comparisons.
And what I'm saying here iswell, maybe the magnitude of the
effect for the other groups waslarge.

(32:02):
So for polarized versus puremiddle, maybe there was zero was
the magnitude, whereas theother ones, you could say maybe
the VO2 max increased by fivemils, but there is such a large
degree of error, of statisticalerror, there that it just didn't
reach a statisticalsignificance.
And I can actually speak tothat likely because the sample
sizes were too large for that.

(32:23):
So you'd see very largestandard deviations and you're
just not going to be able todetermine one if it's actually a
true result.
And then also the direction andthe magnitude of those results.
Right, because the smaller thesample size when you look at a
study, then more room for error.

Speaker 1 (32:38):
Do you think that this is purely an artifact,
though, of the sample size, ordo you think that there are
other things kind of going on asto as why you are not seeing
anything?
At the end of the day, for theother groups.

Speaker 2 (32:48):
Definitely, I think it's sample size there's very
small, there's very fewparticipants.
I think we had some that wasonly like 20 or 30 total
participants in the comparisonsand interestingly, when I
compared between the protocoland the intention to treat for
one of the groups, thecomparison went from one side
all the way to the other.
So it went from favoring oneintervention all the way to the

(33:11):
other and that really suggests asampling error, meaning the
sample size is so small thatmaybe there's three participants
all the way to one side of thedistribution.
Basically, I'm just saying it'sjust too small to really tell
what's going on there so is thetake-home message here I'm
trying to bait you intoanswering some questions.

Speaker 1 (33:28):
Here is the take-home message.
It doesn't matter for anyathlete.
Meaning if you do any sort ofwilly-nilly, you know type of
training structure, highintensity, threshold intensity
or whatever the end result whenwe're looking at time trial,
performance and rvo2,irrespective of the athlete's
experience level or how goodthey are Like, is there any

(33:50):
difference at all between theseand if there is, how can we
start to distill through that?

Speaker 2 (33:57):
So that brings me to, kind of that, my secondary
analysis that I looked at.
So when you see that there's nodifference, you say, okay, well
, am I sure that I combined allthese studies properly?
And so we're looking atsomething called the statistical
heterogeneity.
And so if there is this degreeof variability, you say, well,
what confounding variables caninfluence the results?
Did age influence it?

(34:18):
Did certain participantcharacteristics or training
characteristics?
And so we looked at all ofthese different variables and
specifically with polarizedversus pyramidal, because one,
there was no difference, notthat there was no statistically
significant difference, and thesample size was large enough.
And so this was the onlycomparison where there was
enough participants to lookfurther into the analysis.

(34:43):
And when we looked further intothe analysis, we found that
actually performance levelmatter, which was quite
interesting.
So there was a differencebetween which type of intensity
distribution polarized versuspure middle benefited one type
of athlete versus another.

Speaker 1 (35:00):
And so what was that specifically then?

Speaker 2 (35:02):
Yeah.
So with respect to polarized,we found that competitive
athletes tended to benefit morefrom a polarized distribution,
whereas recreational athletestended to benefit more from a
pure middle model.
Now, when I say tended tobenefit, the reason why I'll be
very particular about my termsis because there wasn't a
statistically significantdifference between those groups.

(35:25):
So it was actually 0.6 standarddeviations.
Basically, it's for that typeof population that's quite large
.
So they really were different.
Not only that, they went incompletely opposite directions.
The only thing was is when Ithen went into looking at these
subgroups, specifically therecreational versus competitive
athletes, the sample sizes startto get smaller, right.
So now we're taking theseinterventions, we're cutting it

(35:47):
in half and now it's evensmaller, and so just the
magnitude of the effect.
So basically, it was about justunder a half the standard
deviation for both groups, butall the way to the other
direction.
So recreational athletes tendedto benefit from polarized and
again I started from pure middleand competitive tended to
benefit from polarized.

Speaker 1 (36:05):
It just didn't quite meet statistical significance
when we're looking at that exactmagnitude there okay, let's
take a quick break in theconversation with michael to
bring in cts coaches ryananderson and adam ferdinandson
to expand on this aspect of howelite athletes might actually
adapt differently to differenttypes of workouts as compared to

(36:27):
their non-elite counterparts.
Okay, ryan, so we have thislike concept from this paper
that elite athletes generally dobetter underneath a polarized
structure.
We'll just say better athletesgenerally do better underneath
the polarized structure andathletes who are not as good do
better underneath a pyramidalstructure.
Let's try to like like breakthis down for what this means

(36:50):
for, like your group of athletes, because you coach athletes
that are both at the elite leveland at the non elite level.
How do you put this piece ofliterature and what Michael and
I have been talking about kindof into context just in in the
day-to-day, from a day-to-daycoaching and training
perspective?

Speaker 4 (37:04):
So one takeaway I have is defining like better
athletes are likely to have moreoverall training history.
It can be kind of aself-selection and that they've
become a better athlete from allthat training history or they
are an athlete who can put inconsistent training to become a

(37:25):
better athlete.
So this was my first thought inthese two different ways to
label a better or experiencedathlete or a less experienced
athlete.
More training history typicallyleads to a better athlete.
More training history means youcan reach a plateau with your
growth.
More training history means youcan reach a plateau with your

(37:47):
growth.
So a phrase I like to use withmy athletes is like okay, what
lever are we going to crank onin training?
And the more training historyyou have, the more experienced.
You kind of have to go to themore extreme end of the spectrum
or the zones or on thatpolarized end of really hard
intensity.
If an athlete is self-coached,I typically find they are less
likely to have hit theseintensities before, maybe out of

(38:10):
fear, or especially ultrarunners like what's the point of
hitting these intensities?
So it is less likely to be anintensity they have trained at
and therefore it's an intensitythey're going to respond better
to.
So that's kind of my thoughts,with better athletes needing

(38:30):
more intense work to respond,and then with the less
experienced athletes, you canjust go with the simple
principle of like, hey, don'tscrew it up, they don't have a
lot of training history yet.
Whatever you give them, they'regoing to keep improving.
They don't have a lot oftraining history, which means
they may not be as durable andable to withstand really hard

(38:52):
workouts.
So don't screw it up.
Give them a little bit ofeverything and they're going to
keep responding and growing.

Speaker 1 (38:59):
I didn't even think about this when I was
interviewing Michael, but thefact that a good athlete with a
lot of training history, thatmight be an elite athlete, it's
a little bit of aself-fulfilling prophecy that
they have done workouts at avariety of intensities because
you don't get to that level, youdon't get to that experience
without having that big varietyof intensities.

(39:21):
Similarly, on the other end ofthe spectrum, most newer
athletes who are inexperienced,don't have that variety of
intensities for whatever reason.
There might be a fear factor,they might not just know how to
do it, they might be runningrecreationally, they want to run
predominantly at a less intenselevel and so in many ways, when
you think about it, like thekind of the results of this
shouldn't be all that surprisingjust because of that

(39:44):
self-fulfilling prophecy, youkind of like get, you kind of
get to this area, or you get tothis level simply as a byproduct
of all the training that you'redoing, adam, I kind of want to
like throw it over to you aswell.
So we've talked about likepolarized and pyramidal training
, kind of in this, with, withthis like overall architecture,
but I was wondering if you couldactually bring that to reality

(40:05):
for the listeners in terms ofwhat does that actually look
like at the workout level?
What sort of workouts would fitinto this construction that
we've been talking about for thelast several minutes?

Speaker 3 (40:14):
Yeah.
So this honestly isn'tsomething that on a day-to-day
basis.
I look at one of my athletes'training plans and say, okay,
we're going to make it morepyramidal this year or more
polarized this year.
It does happen somewhatnaturally from the level of the
athlete, kind of like you weredescribing Coop.
But a more polarized plan wouldprobably include more of the

(40:36):
very high intensity workouts, sowhat we might call VO2 max
workouts or running intervalskind of colloquially here and
then pure middle might include agreater percentage of that
middle range, the tempos and thesteady states, and that's a
really big, big picture, broadbrush way to describe it.
So someone that's more elite,you might spend more time

(40:59):
hitting those workouts, morevolume within those workouts.
So maybe you're talking about afour by four minute workout,
something pretty chunky, even upto like a five by five for the
really really intense workouts,whereas if I have a newer
athlete and we're trying to dosomething that same intensity
range, I might do six by two orby three, you know some sort of

(41:21):
more bite-sized variety likethat.
And it kind of reminds me ofthe principle of don't go there
until you need to go there, andso many of our newer athletes
don't need to go there.
There's a practical andpsychological element as well of
whether or not they canactually execute those workouts
if we gave them bigger ones,whether or not they're mentally

(41:42):
ready to handle that and areused to pushing at those higher
levels of exertion, and I thinkthat takes quite a few years to
get really good at.

Speaker 1 (41:51):
So I'm going to boil it down to like really simple.
So a quote, unquote polarizedworkout in our vocabulary would
be a running interval workoutand that's something like five
by three minutes hard, threeminutes easy.
Seven by three minutes hard,three minutes easy.
For a really good athlete itmight be four by five minutes
hard, two and a half minuteseasy, something of that
construction, where the totalamount of time at intensity is

(42:13):
anywhere between 12 and maybe 20minutes at the very maximum.
A pyramidal structure will alsoinclude those workouts, but
then we'll also include what wecall threshold workouts or maybe
even steady state workouts.
So threshold workouts or temporuns, which would be our
vocabulary, would be somethinglike four by 10 minutes hard,
five minutes easy.
And rather than go through allthe different like permutations

(42:35):
of that, I'm going to drop alink in the show notes to an
interval workout article that Iwrote for the Train Right
website.
The train right Website that youcan look at and you can.
It gives you a menu of any ofthese workouts that we have kind
of like ever prescribed in likea small, medium, large format
based on your actual experiencelevel.

(42:57):
And, interestingly enough, justto give a little bit of a plug
to this, ryan's working on an AIproject with me where we've had
to take the this kind ofconstruction of workouts and
bring it to life to people whoare like applying to this ai
project and make sure that weget them in the right beginner,
intermediate, advancedclassification to make sure that

(43:18):
they have the right volume ofintensity at each one of these
types of each one of these typesof workouts.
We're going to spare thatdiscussion for later as a little
bit of a tease for a podcastthat will probably come out six
weeks from now with the personwho's designing that.
But it's something that verymuch we have to bring to light
day to day from a coachingperspective is not just to look
at this research but also lookat how are we actually going to

(43:40):
program workouts based off ofthis.
So I'm going to likecolloquialize this.
If people are getting lost inthe polarizers pyramidal, the
better athletes benefited morefrom higher intensity
interventions, and the athleteswho weren't as good benefited
more from the what I'll callmore medium intensity or
threshold intensity types ofinterventions.

Speaker 2 (44:01):
Fair categorization there oh, definitely, that's
actually kind of one of thethings that we were considering
is probably why is thishappening?
Like, is it just it had to dowith that intensity?
Yeah.

Speaker 1 (44:10):
Yeah, so.
So that was leads me to my nextquestion why so, if this is a
like a real finding right, why?
Why would this happen in yourmind If you kind of look at this
through your practitioner'slens?
Why would we see differentathletes improve differently or
benefit more I'm going to usethat term benefit more from a

(44:30):
different intervention?

Speaker 2 (44:32):
So, in terms of the analyses that we ran, there was
no difference in terms of thetotal training volume, the
number of weeks, et cetera forany of the studies.
When you know, did thatinfluence those results?
Weeks, et cetera, for any ofthe studies when we're you know,
did that influence thoseresults?
And so the reason why I bringthat up is because a lot of
people would say, well, ifyou're going to do a polarized
training distribution, you'regoing to have to do way more
training volume for that to bebeneficial.

(44:53):
And so what we found was, well,that didn't really matter in
this case, it didn't reallyinfluence the results.
So we, you know, in terms ofour analysis, there wasn't
anything that very specific thatwe could point out other than,
like I said, the performancelevel.
So you say, well, why wouldperformance level influence this
?
And you might think, well, thehigher trained you are, the more

(45:16):
volume of intensity you canhandle, Whereas the less fit you
are, maybe you can't handle asmuch training in zone three.
Either you physically burn outwhile you're trying to do it, or
maybe you're just notrecovering well enough.
It could be any number offactors that we can discuss, but
probably it's that well thesehigher trained athletes handle

(45:39):
more intensity, they recoverbetter and they can generate a
stimulus for change from that,whereas the recreational
athletes can't.
The other way of looking at itcould be well, where is your
relative threshold?
So one way of looking at it iswell, here's your intensities.
And then the other thing isthere's a lot of literature that
shows that recreationalathletes well, maybe their

(45:59):
second threshold is around,let's say, 80%, and maybe- 80%
of their VO2 max.
Sorry, yes, 80% of their VO2 max.
And that maybe competitiveathletes, typically around 90%
on average.
Well, so then where would yousay the weak point is for the
competitive athletes?
Well, it would be pushing theirVO2 max.
And then, so maybe you need todo more higher intensity

(46:22):
training to push your VO2 max.
And then, when you look at therecreational athletes, if
they're doing a little bit morethreshold work, is it?
I mean, this is we're just kindof discussing here.
I can't say this based on mywork, but we could say, well, is
it possible that maybe they'llbenefit from doing more
threshold work to maybe pushthat second threshold a little
bit more, or just not as muchhigh intensity effort, and so

(46:44):
maybe that's where there arelimitations.
We want, maybe we want to shiftthe threshold a little bit and
so that could be another way oflooking at this.

Speaker 1 (46:51):
Well, certainly, because once again I mean having
work, you know, having had aphysiology lab and access to
lots and lots of results, andyou kind of recognize the same
thing.
One of the key differences thatnormally shows up when you see
a whole lot of athletes is thebetter athletes have their
threshold.
However, you want to definethreshold at a higher percent of
their VO2 max, sometimes almostto a fault, to where you've got

(47:14):
to go through different coursecorrections.
Where an athlete's threshold isrelative to their max is
actually one of the bigger keypieces of data that we can
extract from a graded exercisetest, because it tells you and
this is part of this researchright here it tells you what
lever you can push on, whatbutton you can push, and so if

(47:37):
your threshold is really closeto your max, doing maximal,
improving that max is going tohave a bigger benefit.
If your threshold is way, faraway from your max, it's
probably both, and maybe youcould make an argument that the
improving your threshold shouldbe more of the, should be a
bigger part of the emphasis.
You can kind of boil it a lotof ltv02 tests down to just that

(48:01):
when you have a new athlete andif you're trying to extract
information related to what youwant their training program to
actually look like, what youwant their periodized approach
look like is just look at theirpercentage of VO2 max or their
percentage of threshold ascompared to VO2 max and take
cues from there.
Now I'll back up a little bitbefore somebody accuses me of

(48:22):
being myopic or something likethat.
That doesn't mean that all ofthe training goes into that
bucket.
It just means how.
It just means that you take alittle bit of a different
approach and shade more of thetraining to the things that
matter more and shade thetraining away from the things
that matter the least.

(48:42):
When you look at it over alonger period of time so nine
months, 12 months, 24 months andthings like that If you know an
athlete is quote unquote weakin one area, you just design
everything so you have a.
You have more of the trainingto accentuate that.

Speaker 2 (48:55):
No, exactly, and that you know, it's exactly the way
that I think it should be doneas well, and it's I like how you
kind of said well, it's justkind of shifting a little bit of
a focus more into where theirlimitations or where their
weakness is, and I think that'svery important.
In fact, you know, hopefullywe're going to be analyzing some
data to look at.
There's a study that came out,a 2022 study.

(49:16):
It's actually one of thestudies that we included in our
analysis that actuallyperiodized, pyramidal and
polarized training, and so whatwe're going to do is actually
look at responders versusnon-responders and look at their
relative thresholds and see,well, maybe there are some
responders that benefited morebased on where their threshold

(49:39):
was.
So, for instance, if they hadvery high second thresholds,
maybe they benefited more from apolarized model to push the
ceiling or push the roof,whereas maybe those who tended
to have lower thresholdbenefited more from a pure metal
model.
And then, when they in terms ofthe periodized approach, and so
we're going to, we're going tokind of reanalyze our data to

(50:00):
see if if that may actually showsomething there.

Speaker 1 (50:02):
Yeah, yeah, and I mean I think this is a good
pause point to kind of moveoutside of the research and into
some of the more practicalelements that you just mentioned
.
It would be very rare and I'mspeculating a little bit on this
, but I don't think that that'stoo far of a speculation it
would be very rare that you hadan endurance athlete with any

(50:23):
sort of reasonable history two,three, four or five years of
training or something like thatwas only using one intensity
domain for their intensity whenyou looked at it over long
periods of time.
Most of the time they're doingzone three and zone two and zone
one and zone back to zone twoand zone three when you look at
it, sometimes within a week,sometimes with even with even a

(50:45):
session like.
So that's the way some, youknow, interval sessions would
kind of be designed, and so Ithink that when we start talking
about this, we have to realizethat it's not a going back to my
kind of my first point.
It's not a versus proposition,it's when you roll it into the,
into the practical elements.
It's it's how much proportionare you spending at these

(51:07):
intensities versus theseintensities?
Is it 50-50?
, is it 70-30?
And I definitely think that youcan take some shades from that.
You can take some shades ofthat with this particular
research and we have history tokind of like back this up as
well as well that generallyspeaking, athletes with a longer
training history and that arequote unquote better are going

(51:28):
to benefit more from higherintensity work and less from
threshold work.
And then the opposite mightalso be true with athletes that
are a little bit earlier intheir whole endurance journey,
like if you want to take thebiggest, broadest brushstroke
from a coaching standpoint forthis type of research.
That's what I come away with.
It's just like what intensitydistribution would I use over a

(51:50):
12-month period with any type ofathlete?
And if they're more experienced, I'm going to shade that
intensity distribution towardsmore higher intensity stuff, and
if it's more of an entry-levelathlete, we're going to spread
it out a little bit more.

Speaker 2 (52:03):
Yeah, because I mean, if you're a highly trained
athlete, you may respond muchquicker, so you may be able to
put more time into somethingelse, you may recover quicker
and it changes right.
It's like you know, like wesaid earlier, how we're
comparing training intensity,distribution, interventions, and
so it's not just whichintervention is better, but when

(52:24):
you go down into thecharacteristics of the
individual, why might oneindividual benefit from one
distribution over another?
And then you say, well, nowthat they've adapted, are they
now a different individual anddo they now subsequently need?
a different model, and so it'sjust what do you need right now?
And then, as you adapt, where'syour limitation?

(52:48):
And then let's address thatlimitation.

Speaker 1 (52:50):
Yeah, and that's actually where testing can
really come into benefit,because you see the effects of
the training, but then you alsosee if the limiter that you
previously thought was there haschanged and if there is a
different limiter that you nowneed to address.
So, like wonderful wrap upthere.
Okay, let's take another quickbreak in the conversation to

(53:11):
bring back CTS coaches RyanAnderson and Adam Ferdinandson
to specifically discuss how weuse physiological testing to
further individualize thetraining that we are giving to
our athletes.
Individualize the training thatwe are giving to our athletes.
So, ryan, you've actually hadathletes come into our

(53:33):
physiology lab, do physiologicaltesting and then, based off of
the results of thatphysiological testing, had to
decide what to do with theirtraining that might be different
from what you originallyplanned.
Why don't you describe thatprocess?

Speaker 4 (53:43):
Okay.
So this athlete had a robusttraining history, which was also
in training peaks, which isalways so nice when we can look
at their data in training peaksand use our ways of finding
specific intensities, workoutsat those intensities, et cetera.
So this athlete's traininghistory is robust.

(54:04):
They put in a lot of volume andnaturally, from onboarding this
athlete and talking with them,it was like, yeah, they did a
lot of volume, not super highintensity.
I'm not seeing a lot of truelike VO2 max running interval
workouts.
I am predicting that is notgoing to be their strength.
That wasn't that hard toconclude.

(54:25):
They go in the lab, sure enough.
Their VO2 max is a relativelylow percentage of their
threshold, but the magnitudethat I saw that at was like, oh
crap, this is an even biggerthing that we need to leverage.
This athlete's goal events are100K, 100 mile, so we want to be

(54:52):
specific to those workendurance, bigger volume that
they can handle.
But seeing that datareemphasized, to me it is well
worth this athlete's developmentand getting most prepared for
these events by doing anadditional running interval
block that I would haveoriginally planned.

(55:14):
So in this case the testingshowed what was easy to
hypothesize and be correct, butthe magnitude and just seeing
the data was a reminder that,while this intensity is not most
specific to these athletes goalevents over the next six to
eight months it is going togreatly benefit them and help

(55:36):
their development so you're justto like encapsulate that a
little bit.

Speaker 1 (55:40):
Their threshold compared to their max was, let's
just say, 70 or something likethat.
We would consider that farbelow their max.
And then what you did as aconsequence of that is you
shaded a little bit more of thetraining towards that specific
intensity in order to push it alittle bit closer to the max
exactly and we.
So this is where testing testingcan become actually a better

(56:06):
tool than actually looking atthe training history, or it is
something that makes you doubledown on what you were going to
do earlier based off of thetraining history.
Most of the time it's the latter, because most of the time what
we see in the physiology lab isa reflection of what the athlete
has actually done.
Lab is a reflection of what theathlete has actually done, and
if you have very good training,if it's all logged in training

(56:29):
peaks and they've got good notesand it's three or four years
worth of data sometimes, there'smost of the time there's not
any sort of like big surprisesright, you got what you got,
it's been revealed through thetraining process and you're just
getting a little bit of a moreprecise lens on it through
through through actualphysiological testing.
But having the data as opposedto your opinion of the training
process, it becomes one of thosethings that kind of engenders

(56:52):
some further confidence,especially for a new athlete,
because now you have twodirectional arrows that are
pointing the same way.
One is an analysis of thetraining, the other one is an
analysis of the athlete'scurrent physiology, a snapshot
of their physiology, and whenyou have both of those kind of
pointing in the same direction,it gives you and the athlete a
lot of confidence for whatyou're actually doing.

(57:13):
I'm going to try to put somenumbers onto this at the very
end of this quick conversation,just so athletes who have
physiological testing can kindof bring it to light or turn it
into reality.
But, adam, I'm wondering, fromyour perspective, what do you
have to say to this concept ofusing physiological testing to
start to shade the training,either based off of Ryan's

(57:34):
example or of any anything thatyou've kind of encountered?

Speaker 3 (57:36):
Yeah, well, in general, it would be very nice
to have physiological testingfor all of my athletes.
I have physiological testingfor two and one of them was, I
don't think, done very well,went over that with someone that
works here internally.
And then the other one it waslactate only.
It just it kind of checks out.
It looked good.
We're going to continue ourtraining.

(57:58):
So these examples where we see areally clear directional arrow
are really nice.
But in lieu of that, and evenI'm sure if we did more testing,
the magnitude of thatdirectional arrow probably
doesn't outweigh a lot of theother factors that we're
balancing Whether the athleteneeds to go work on technical
terrain, more fueling oranything like that.

(58:20):
I think the situations where itleads to a clear decision on
training are kind of few and farbetween, at least for me
personally and the athletes thatI work with.
So I make a lot more decisionsbased off those kind of former
things.
I mentioned the more practicalparts and you know, occasionally
without testing, you can seesomething pretty clearly in

(58:40):
training peaks and in their data, but not as often as I guess as
I would like.

Speaker 1 (58:45):
Okay, so really good point here.
So a big piece of Michael'sstudy.
It was digesting training dataand that was that's what he
spent an inordinate amount oftime actually doing with the
subject pool.
And one of the reasons that'sso important as a coach trying
to get a handle on a new athleteor actually what is going on
with one of your existingathletes is that you have many

(59:06):
data points to pull from.
Not all those data points aregoing to agree with each other,
but you have a lot of them.
One of the disadvantages tophysiological testing it's a
data point from a singularmoment in time.
It's a reflection of thatathlete's physiology, their
capacity at that one moment,which can change.

(59:27):
And most of the time when we'reworking as coaches with
athletes and we have testingdata and training data, the
training data is the biggerdirectional arrow.
So if we have these two thingsthat are both pointing in the
same direction, it's great.
But if you have these twothings that are maybe, you know,
10 degrees off or somethinglike that, more often than not

(59:48):
we're looking at the trainingdata to try to come to some sort
of confluence as what to do, asopposed to the testing data,
which a lot of people actuallysee is a little bit backwards,
right, because the test issupposed to be accurate and
you're measuring all thesethings.
You're measuring oxygenconsumption, you have lactate
draws and there's a lot ofpeople involved in it.
It's usually a big effort to goout to a lab and things like

(01:00:08):
that, but I think what a lot ofpeople have to realize is that
the scope of that actually isrelatively limited, both in
terms of what you actually dowith it and then and then and
also how much you actually, howmuch you actually rely on it.
So, to your point of gettingthe right testing data, we could
adjudicate that for anotherwhole podcast and go through

(01:00:30):
what labs to look through whatlabs to look at how to kind of
discern if you're going to getgood lab data or bad lab data.
I'm not going to go throughthat right now, but if you do
have, if you do have good labdata, I think one of the first
things to look at that'srelevant to this conversation is
what is your lactate thresholdcompared to your VO2 max?

(01:00:53):
This is kind of one of the keynumbers to determine training
architecture and we've said thatif it's quote unquote low, that
you can take this type ofapproach.
And if it's quote unquote high,then you can take this type of
approach.
And if it's quote unquote high,then you can take that type of
approach.
I almost like to use like anexclusionary rule set when I'm
looking at this.
So if it's not high, it's kindof medium.

(01:01:13):
So if you get somebody who'slactate threshold is within 95
or maybe even 99% of their VO2max, they don't have a lot of
room to run by doing a lot ofthreshold intervals because
they're so close to their maxalready.
So therefore you would take ona more of a polarized shade to

(01:01:33):
their training and improve theirVO2 max first.
I think everybody else fitsinto the other category, meaning
they can benefit from a widevariety of training and you can
kind of take the trainingarchitecture wherever you want
to.
There are certain cases wherethat threshold is so low lower
than 75% of their VO2 max whereyou would do a disproportionate

(01:01:57):
amount of low intensity work,steady state work, tempo work
and things like that, in orderto bring that up before you even
touch VO2 max work.
But those are, I would say,kind of more rare.
So for those of you that aretrying to use lab testing to
determine any of this, I wouldfirst encourage you to look at
your training history, becausethat's going to be a better

(01:02:19):
directional error.
And just what types of traininghave you done?
Have you done adisproportionate amount of low
intensity or threshold work?
Well, if you've done adisproportionate amount of that,
you should change it up alittle bit in order to get kind
of a further, kind of furtheradaptation.

(01:02:45):
I've had even elite athletes goto labs and I've been kind of
ridiculously underwhelmed withwhat they have come back with.
It's not, it's not.
It's kind of not an easyproposition, but go find a high
quality lab and see if those twosets of information actually
come together in the same andare telling you the same thing.
I need to do this type of workbased off of the history and I
need to do this type of workbased off of the, based off of
the physiology, and I need youto do this type of work based

(01:03:07):
off of the physiology.
And if those two things don'tcome together, go get an expert
opinion on it.
That's usually not the peoplein the lab, that's usually a
coach or a practitioner who'sactually had to synthesize all
this information in order toknow what to do.
The labs are really great attelling you what your numbers
are and telling you where, whatyour ranges are, what they, what
many of them they're.

(01:03:29):
The failure point of many ofthem is then being able to take
that and put it into the kind ofthe real world, because they
don't have your training historyin a lot of cases and it's a
big lift to look at all thattraining history.
So don't think that you'regetting like Baghdad or a bad
deal or anything like that.
It's just a big lift to dothose things.
But marry those two things up.
Marry the training piece of itup with the testing piece.

(01:03:52):
Don't overly rely on one or theother, and I think that is more
of a surefire way to get a gooddirection on what type of
training you want to deploy forthe next three, six, nine months
, however long you're likelaying it out for.
Okay, let's get to audiencespecific now.
So this is an ultra marathonaudience and pretty much only an

(01:04:12):
ultra marathon audience.
I mean, maybe less than 5% ofmy audience is going to come
from Ultraman and Ironman andcycling and things like that.
This is really specific.
And cycling and things likethat this is really specific.
And whenever we talk abouttraining studies that either are
discussing high intensity,polarized training via two max
types of intervals and thingslike that, and or research

(01:04:35):
outcomes that are based on atime trial, some people will
just roll their eyes in theirback of the, in the back of
their head, saying okay, what?
Why is this relevant to me whenmy event is six hours, 16 hours
, 24 hours, 30 hours?
Why should I be looking atthese?
Why should I be taking cuesfrom this research that that, on

(01:04:56):
appearance, is in a differentdomain with a different time
specificity?
I'm wondering what you have tosay to that just specific to the
audience, and how we can makeit relevant for them.

Speaker 2 (01:05:06):
I think that's a very important question and,
interestingly, I don't thinkit's just an ultra endurance
sport problem now you know, andI say to them I mean cyclists
would say, well, we can't trainthe way runners do, or runners
will say, we can't train the waycyclists do, and regardless of
the distance.
So what I've kind of learnedover the last however many years

(01:05:27):
is that, okay, well, we, youhave a certain goal and you need
to think of, well, what's thegoal that you have?
And in this case, if it's ultraendurance, and you say, well,
what are the physiologicaladaptations that I need to occur
to then be able to achieve thatgoal?
Now there's a whole othercomponent to add on to this with
ultra endurance athletes andI've coached several Ironman

(01:05:49):
athletes, and so that's about asclose as I can get to thinking
about ultra endurance and that'swhere nutrition really starts
coming in.
And then there's these otherfactors that I certainly am
nowhere close to being aspecialist, both on the science
side as well as on the coachside, to get into.
But what I would say is, if youneed to increase your VO2 max,

(01:06:10):
then you need to do a trainingprogram specific to increase
your VO2 max.
If you need to increase yoursub-maximal thresholds be it
your LT1 or your LT2, then youneed to do an intervention
that's specific to doing that,and it doesn't matter who you
are.
You could be a 40K runner oryou can be a 10K runner or an
ultra Pushing.

(01:06:31):
That first threshold is thesame type of training regardless
.
The interesting thing withsomething like ultra endurance
is well, what's the need in thesport?
So 10k runner would need tohave need to be able to push at
much higher intensities duringtheir race, whereas somebody
who's an ultra endurance atultra endurance athlete is going

(01:06:52):
to be spending a lot more timeat a lower zone.
So it's not necessarily themethod to achieve that specific
physiological adaptation.
It's going to be the same, nomatter who you are.
It's just where.
Where should you focus?
And that's what I think is whatit comes down to.
Somebody was saying to me oh,should I do one-minute intervals
, or should I do four-minuteintervals, or should I do
six-minute intervals?

(01:07:13):
And now I'm generalizing bigtime, and I said this in our
last discussion.
Well, maybe six-minuteintervals are the optimal way to
achieve that physiologicaladaptation, assuming you're all
the same level of training andthe same experience.

Speaker 1 (01:07:30):
So then, why would you do four minute intervals if
six minutes is the best Right?
Always optimize, alwaysoptimize.

Speaker 2 (01:07:32):
Right.
So you do the thing that'sgoing to lead to the best.
I was saying, if you've neverdone an interval session before,
one minute's going to kick yourbutt just enough.
You don't need to do six minuteintervals, Right?
But the point is, is what's thebest way to achieve that
physiological adaptation?
And then that's the program todo.
But the question is again goingback to ultra endurance.
Well, what's the need?

(01:07:53):
What ultra endurance athletesneed, and they need to be able
to really push that firstthreshold as far as they can,
which involves a huge amount offat oxidation.
And then there's a lot ofdietary components to that too.
So I'm not getting.
The thing is, I guess I'm notgiving the best answer to it,
but what I'm doing is changinghow we think about training.
It's not that training isspecific for an athlete.

(01:08:14):
It's well, what should I interms of the physiological
approach?
It's well, what's your goal?
And then hit that.

Speaker 1 (01:08:19):
That that method to get to that goal, right?
I mean, simply put, you canimprove the physiology that's
still relevant to the event,without with using intensity
domains that are seeminglyunrelated.
That's the way I think about itis okay, yeah, you're doing a
24 hour event, right?
You're well below your firstlactate threshold, your first

(01:08:39):
threshold, ventilatory threshold, however you want to define it
for the entirety of that event.
That doesn't mean that the onlyrelative, the only way to
improve is work below the firstthreshold.
Sure, that's a big component ofit, because that might be the
most important thing, but thatbecomes really limiting at the
end of the day.
And if you're only focusing onone thing, because we talk about

(01:09:00):
this concept of having a stringof an athlete that can operate
across a lot of differentintensities, and if you push or
pull on one, one part of thestring, that whole string moves,
so, so, so anyway, I mean thatwhen I kind of come back to,
when I try to answer thequestion of why are these
intensities, they're seeminglyso irrelevant to the event that
I'm doing.
Why am I doing these?

(01:09:21):
I just I come back to that asis.
You'll still improve and you'llprobably improve more at the
intensity that you have toactually operate at, because
there are physiologicaladaptations that you are going
to accentuate only at thoseintensities, or maybe
predominantly those intensitiesthat you can't at other,
intensities that make theentirety of the athlete better.

Speaker 2 (01:09:41):
Yeah, we think about if you train your VO2 max.
You're raising the roof.
You're raising the roof, butyou're raising the roof, You're
also going to pull the secondand first, or a long Yep, a
hundred percent, a hundredpercent.

Speaker 1 (01:09:51):
So what else I mean?
What else can athletes andcoaches kind of like take away
from this from a programmingstandpoint?
So this podcast is going tocome out If I, if I get my stuff
together.
I've had a little bit of apodcast lull for the people that
are in the know, and this isgoing to be one of the first few
that I release after that lull.
It'll be released in February.
People are hopefully juststarting to execute their plans

(01:10:16):
right.
They have their seasons kind ofmapped out in front of them.
I know what my apex race is.
I know what the training racesthat I need to do in advance of
that are.
I've got my next six to ninemonths mapped out in terms of a
racing perspective.
So studies like this andthinking about what they mean
are incredibly important duringthis time of year because you're
trying to plan out thearchitecture for the entire of

(01:10:37):
the season.
What from the study can we kindof like use in that process to
like really drill it home foreither the coaches that are
listening or the athletes thatare kind of programming their
own training when they'rethinking about the season, when
they're thinking about theseason as a whole.
So I think using.

Speaker 2 (01:10:54):
It's interesting.
It brings us back to the ideaor concept of reverse
periodization we're startingwith.
Yeah, I know you're talkingabout types of periodization I
love the vocabulary.

Speaker 1 (01:11:03):
Right like periodization is this and the
opposite is reverseperiodization yeah, I know, but,
uh, I know.

Speaker 2 (01:11:08):
But so really what it comes down to is the results
show okay, lower intensity isbetter for recreational athletes
.
So maybe if you're starting offyour season, it might be better
to start with a little bit lessintensity, more low intensity
and threshold work, and thenmaybe possibly shift into
increasing the intensity as theprogram, as an athlete's program

(01:11:30):
, goes on.
Sounds like something we allalready know and have discussed
so many times, but we are seeinga very big difference between
recreational, competitiveathletes, and it does have a lot
to do with their trainingstatus and so and their fitness
in terms of their VO2.
So certainly that would beimportant to consider.
One thing that we didn't discussthat I will say that the

(01:11:52):
results of the study showedindirectly was there was an
improvement in VO2 maxes, whichis what we're talking about.
There was no difference acrossinterventions and time trial
performance, and so the reasonwhy I'm bringing that up now is
because the point is that it'snot just going to be one method
that's going to improve.
You have to look at anathlete's whole program and make

(01:12:13):
sure that you're modifying thatprogram along the way as
appropriate for that individualathlete, by constantly
reassessing as appropriate andthen shifting to address their
weaknesses assuming it's likelytheir weaknesses to then to see
those changes.
If we stay consistent and justsay, well, we're just going to
do a pure polarized model andjust keep it exactly the same,

(01:12:37):
then we may not necessarily seethe adaptations we're hoping for
in overall racing performance,and so it was kind of an
indirect result of the studythat we saw.
So it's again we're all science, just kind of looks at an
intervention.

Speaker 1 (01:12:51):
It's hard to look at the whole program as a whole,
and that's where the coachreally needs to come in yeah, I
like that takeaway though,because a lot of times you know
you're going to combine I mean,you're combining a lot of
different studies.
But a lot of times we'recombining a lot of different
studies as well as what we'veseen in the field to determine
what to do with athletes and inin my mind not to use too much
like confirmation bias oranything like that that this

(01:13:14):
kind of validates thearchitecture that we use a lot
of the time with athletes whenwe're looking at how to plan an
entire season out.
We want to do a variety ofdifferent intensities for some
athletes.
The better athletes we're goingto how to plan an entire season
out, we want to do a variety ofdifferent intensities for some
athletes.
The better athletes we're goingto use, you know, a bigger
distribution of the VO two maxwork for the more for athletes
that are a little bit earlier intheir journey or who are not as
competitive.

(01:13:34):
We're going to use more of akind of more of a mixed model.
I think that all of thosethings, we, many coaches,
including myself I won't say allcoaches, but many coaches,
including myself we're alreadykind of like taking that
underneath our wings to to putinto practice.
And this just makes us look atthat and go yeah, okay.
And if you're not doing that, Ido think you should like take a

(01:13:55):
step back and go okay, if I'monly using this type of model,
or only using that type of model, or using a disproportionate
amount of, of of polarizedtraining for an entry-level
athlete, take a little bit of astep back and like think if
that's like the best thing,based on what everything is,
what everything, what all theresearch results are actually
telling you.

Speaker 2 (01:14:15):
Yeah, and then the thing to know when to change is
is to test your athletes.

Speaker 4 (01:14:19):
We all like to say here do four weeks of this, and
now do four weeks of that.

Speaker 2 (01:14:22):
If you don't test your athlete, maybe they needed
to change sooner, maybe theyneeded to change what they're
doing later they're able to keepbenefiting.

Speaker 1 (01:14:32):
So testing is it's actually more important than we
think it is testing and look,I'll actually like expand upon
that a little bit, testing andactually really scrutinizing the
workouts, because you can get alot of value, just you can get
a lot of value, just.
You can get a lot of value thatyou are also looking to get out
of with a test.
If you scrutinize and you haveto do this with a fairly fine
tooth comb if you scrutinize theworkouts, you can pick up on

(01:14:54):
some of those little subtlechanges in where they're
performing at certain parts ofcertain parts of the intensity.
Cycling, that's not all thathard to do.
Running, it's a little bitharder, trail running it's.
I would say it's kind ofobnoxiously hard to do, but
still you can.
But still you can do it.
And I think when you combinethat with some type of testing,
whether it's laboratory, fieldtesting or a combination of both

(01:15:15):
of them, you can really get agood picture with what's moving
the needle and what's not.
Yeah, and when it's time tochange, exactly when it's time
to change, that's the when it'stime to change, that's the best
thing, right?
Because then you know thatyou're not.
You're not kind of like beatinga dead horse, so to speak.
We're going to let you go, man.
This is great, as always.
Like I said, things will be inthe show notes to this piece of
research, as well as theprevious podcast that we did, as

(01:15:41):
well as all fascinating stuff.
I appreciate you coming on thepodcast again.
Where can people get to know alittle bit more about you, the
work you do and the coachingthat you do?

Speaker 2 (01:15:50):
So you can find me, my research and my website is
evidencebasedcoachingca, and I'malso affiliated with the Sylvan
Adams Sport Institute, and soI'm doing some research with
them and hopefully you can beputting out some cool, some cool
work in the next little whileas well awesome and is the
research going to be open access, or is this something where
they're going to have to reachout to you to go and get?

(01:16:12):
so this just I actually that'sfunny that you say that I tried
to get this to be open access,because that's what I really go
for.
But I wasn't able to I know Iwasn't able to make this open
access, while I actually thedelayed in the in the
publication itself was that waspart of the reason why I was
pushing for open access.

Speaker 1 (01:16:26):
Okay, all right.
No, fair, fair enough.
I know this is a constantsource of consternation amongst
the field and we won't belaborthat point here.
News to say there's, there areways you can get it yes okay,
perfect.
Thanks for coming on thepodcast.

Speaker 2 (01:16:39):
I really appreciate it again yeah, no, thanks for
having me again.

Speaker 1 (01:16:44):
I always appreciate his research because it tends to
blend what we see in thescientific community with what
he actually does with athleteson the ground, and I do think
there are things that come aboutfrom his research that you can
apply directly into yourtraining.
We are now getting into theracing season.
We should be thinking aboutwhat types of workouts we are
doing to help tune our bodiesthe best for the races that are

(01:17:08):
in front of us.
Also, thanks to CTS coachesRyan Anderson and Adam
Ferdinandson for offering theirparticular perspective, as well
as CTS coaches FredSabatore-Pastor and Adam for
their production assistance withthis podcast and doing some of
the background research.
All right, folks, that is itfor today.
As always, this podcast isbrought to you advertisement and

(01:17:30):
sponsorship free.
Once again, no ads, no sponsorson this podcast, and that is so
we can deliver the mostup-to-date, unadulterated,
unfiltered information to you,and we do have a heap of that
upcoming.
Since we have had a break onthis podcast for a while, I can
kind of preview some of thepodcasts that I have in the can

(01:17:51):
that are going to be coming outover the next few weeks.
We're going to be talking aboutcarbohydrate consumption some
of the dietary manipulationsthat you can make in order to
potentially improve your raceday nutrition performance.
We're also going to have manyconversations with our coaches
about how to implement thingslike strength training and other
points of training designpractical elements that you can

(01:18:14):
actually take into your trainingtomorrow.
All of these are queued up overthe course of the next few
weeks and we're back.
We're back on a regularschedule.
Expect this podcast to bereleased weekly from here on out
.
Like I said, I'm really excitedabout the content that we have
banked and that is coming up,and as well as the format where
we get to hear from our own CTScoaches to bring to light some

(01:18:37):
of the things that we talk aboutduring our main guest
conversation.
As always, appreciate the heckout of all you listeners.
This podcast is nothing withoutyou.
Appreciate those who havereached out to me during the
course of this break of thepodcast encouraging me to get it
back online.
It has been a long time coming,but those words of
encouragement have meant a lotto me and getting this thing

(01:18:58):
booted back up.
But here we are, we're backlive and we are not going to
stop.
So, like I said, we will seeyou guys next week, appreciate
the heck out of each and everyone of you and, as always, we'll
see you all out on the trails.

All Episodes

Episode Transcript

Popular Podcasts

Stuff You Should Know

Dateline NBC

On Purpose with Jay Shetty

.css-15opob5{left:0;position:absolute;top:0.8rem;} All Episodes

.css-14f5ked{margin:0;word-break:break-word;display:-webkit-box;-webkit-box-orient:vertical;box-orient:vertical;-webkit-line-clamp:2;overflow:hidden;}Polarized or Pyramidal Training for Ultrarunning with Michael Rosenblat, PhD #238

Episode Transcript

Popular Podcasts

.css-r6mb8g{margin:0;word-break:break-word;display:-webkit-box;-webkit-box-orient:vertical;box-orient:vertical;-webkit-line-clamp:1;overflow:hidden;}Stuff You Should Know

Dateline NBC

On Purpose with Jay Shetty

All Episodes

Polarized or Pyramidal Training for Ultrarunning with Michael Rosenblat, PhD #238

Stuff You Should Know